1

THE SYNTAX OF SENTENTIAL COMPLEMENTS

INF SHONA

byNorris. Clemens DEMBETEMBE

Thesis presented to the University- of London for the degree- ofDoctor of Philosophy

School of Oriental and African Studies

1976

ProQuest Number: 10672854

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a com p le te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

uestProQuest 10672854

Published by ProQuest LLC(2017). Copyright of the Dissertation is held by the Author.

All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode

Microform Edition © ProQuest LLC.

ProQuest LLC.789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, Ml 48106- 1346

2

ABSTRACT

This study is concerned with embedded sentences in Shona which are not. relative clauses. These complements seem to he of two types, namely, those dominated by BP which are referred to as noun phrase complements and those directly dominated by VP which are referred

to as verb phrase complements. The case for each type is argued in the first chapter. The second chapter deals with the complementizing mor­phemes found in Shona (Zezuru dialect), A procedure is adopted by which, compiementizers are identified. This study also brings out the fact that infinitival complements and derived nomimals should be treated differently. Evidence is adduced to show that the former are derived transformationally, but not the latter. Chapter Four discusses the in­sertion of complementizers into the underlying structure by analyzing first the transformational approach, then its inadequacy, and finally the phrase structure approach. The latter approach is preferred to the former on empirical evidence. The transformational rules which are mentioned in this study are described in the fifth chapter. Although some of the rules are dealt with fairly extensively, for others no more than an outline is given. Th^main purpose of this chapter was to provide only enough information about how these rules work in order to facilit­

ate understanding of the thesis. The grammatical function of embedded sentences, that is, whether a given complement clause functions as a subject or an object etc, is the concern of Chapter Six.

I consider the main contribution of this thesis to be the chapters dealing with (a) the arguments for setting up the two types

of sentence complements, (b) procedures for the discovery of compleme­ntizers, (c) the arguments showing how complementizers should be inser­

ted into the underlying structure, and (d) the distribution in sentences

of embedded clauses.

ACKNOWLEDGEMENTS

My gratitude goes to the Association of Commonwealth

Universities for the Commonwealth. Scolarship Award, which enabled me to do the greater part of the research for this thesis at the University of London*.

I should like also to acknowledge my indebtedness to my supervisors, Dr.J.H.Carter of the School of Oriental and African Studies and Dr.il.V.,Smith of the: University College London., and also to Professor G.Fortune, University of Rhodesia, for their valuable criticisms, suggestions and encouragement.

Salisbury February, 1976 N.C.D.

4

CONTENTSpage

TITLE 1ABSTRACT 2

ACKNOWLEDGEMENTS 3CONTENTS 4INTRO DU CTIONT 8

1.0 SENTENTIAL COMPLEMENTS OF VERBS 121*1 General 121..2 Tie' Case For* Noun Phrase. Complements 121*2*1 Tests for noun phrases in subject position 12

1..2*1*1 Comcordial agreement 13

l.*2*l*2 Interrogative pro™forms 141*2*2 Tests for noun phrases in object position 161*2*2*1 Passive test 161*2*.2*2 Interrogative pro-form test 21

1*2*2*3 Object anaphora test 231*2*2*!:. Pseudo-eleft test 231 *2 *3 NToum phrase complements 27

1*3 The Case For Verb Phrase Complements 411*4 Sentential Complements and Seleetional Restrictions 46

2*0 COMPLEMENTIZING MORPHEMES 31

2*1 Kuti- /+ indicative/ end kuti /+ subjunctive/ 512*2. Ku /+ infinitive/ 522*3 Kuti 60

2*3*1 Kuti. (to say) 652*3*2 Kuti (to bring about.) 702*3*3 Kuti (a complementizer) 742*3*4 Kunzi 83

2*4 Kara /+ consecutive/ 86

2.5 Kama /+ participial/

5pag8?

2.5*1 Direct questions 88

2 *5 *2 Kama followed by indirect questions 892 .6 Deletion of kana 912 .6 .1 Kana (time when) 912 * 6'.* 2 Kana (if) 93

2*6.3 Kana (although, even, if) 94

2 . 6 *4 ICana (whether) 95

2*7 Zva /+ participial/ 95

2 .8 Deleted Complementizer Formatives 97

2.9 Verb, -ti (say) and Complementizers 992..10 List of Complementizers 100

2*11 Kuti /+ subjunctive/ and Ku /+ infinitive/ 101

2 *1 2 Kuti /+ indicative/ and Ku /+ infinitive/ 1072.13 Summary 109

3.0 INFINITIVAL AND DERIVED NOMINALS 111

3.1 Productivity 1133.2 Meaning 116

3.3 Internal Structure. 117

4 *0 COMPLEMENTIZERS AND THE UNDERLYING STRUCTURE 1274.1 The Transformational Hypothesis 1274.2 Inadequacies of the Transformational Hypothesis 140

4.3 The Phrase Structure. Hypothesis 1514.4 Complementizers and Non-embedded Sentences 1744.5 Summary 175

5.0 TRANSFORMATIONAL RULES 1765.1 Passive 176

5.2 Pseudo-passive 185

5.3 Agent Deletion 188

6

page.3*1|i Reflexive 1903.3 Mood Incorporation 1923.6 Complementizer Placement 199

3*7' Gender Copying 198

9*8 Object Raising 2025*9 Subject. Raising 2119.10 Equi-NP Deletion 216

9*11 Extraposition 2199.12 Pro Deletion 2273*13 Object Anaphora 2289*14 Object Focussing 2323.13 Object Copying 2333.16. Noun Prefix (segment) 2363*17 Subject Postposing 2373*18 Rule Ordering and the Notion of the Cycle 2383.18*1 Passive and Object Raising rules 2393*18*2 Reflexive and Subject Raising rules 299

3*18*3 Gender Copying and Subject Raising rules 299

3*18*9 Summary 231

6*0 THE GRAMMATICAL FUNCTION OF COMPLEMENT CLAUSES 2336.1 Noun Phrase Complementation 2336*1.1 Complement clauses in subject position 2336*1*2 Complement clauses as direct object 237'6.1.2.1 Sentences with direct objects} but not

indirect objects 239

6 .1.2*2 Sentences with both direct and indirect objects 2616.1.3 Complement clauses as object of /ir/ 262

6*1*9 Complement clauses as object of /na/ 2?2

6.1.3 Complement clauses introduced by kana /+ partic.7

7

pageand gva /+ participial/ 278

6*1*6 Complement clauses in both subject and objectpositions 280

6*1*6*.1 Verbs of "meaning" 2806*1*6.2 Verbs of "causation" 283

6*1 *6*3 Distributional chart of compLemientisers 286

6*2 Verb Phrase Complementation 2876*2*1 Verb phrase complements of the first type 288

6*2*1.1 Modal-like verbs 2896 *2*1*2 Contracted infinitives 2906*2*2 Verb phrase complements of the second type 292

6*3 Summary 293

7.0 COME RESIDUAL PROBLEMS 2967*1 Relationship between verbs and ideophones 2967*2 An aspect of the possessive construction 3007*3 Complement clauses introduced by kuti /+ subjunctive/

and hu. /+ infinitive/ 301

8*0 CONCLUSION 30k

APPENDIX A 307APPENDIX B 309

BIBLIOGRAPHY 313

8

INTRODUCTION

This s t u d y is an attempt to describe complement or embeddedsentences in Shona. In part it is a reaction to the treatmentof complement sentences in the constituent structure model as

1reflected in Shona Grammatical Constructions. In that work there were some problems which were encountered and which remained unsolved. Some of them are outlined below.

In the work referred to above some complement sentences

are treated as part of auxiliary verb phrases which are defined

structurally as:+ -auxiliary verb radical + complement

This complement is in fact an embedded clause. What is implied in this structural definition is that the auxiliary verb radical and its complement are two immediate constituents of this intermediate structure. The auxiliary verb phrase is then said to be inflected for one or another of the moods in Shona. The inflected auxiliary verb phrase is difined as:

+ ■inflection + auxiliary verb phrase implying that the infledtion and the auxiliary verb phrase are

two immediate constituents.E.g. (l) Ndi^ri kutora mari. 2 (I am taking some money)

2(2) Ha-ndi-na kutora mari (l did not take any money)

4

(1) G. Fortune in Shona Grammatical Constructions, Part II (forthcoming).

(2) These examples are taken from Shona Grammatical Constructionsp. 107.

(3)/ / /Ku-nge ndatora mari

3

(my having taken money)

4(The horizontal line between 1 and 2 in each example represents

the auxiliary verb phrase, while the line between 3 and 4represents the inflected auxiliary verb phrase.)

Although this analysis correctly informs us that thereare two verbs in each of these sentences above, it however fails

/ /to tell us what the grammatical subject of the verb kutora is

in (1) and (2) and of kunge in (3)> nor are told what hashappened to these subjects in the course of the derivation. If

our grammar is going to be adequate all these facts must be described The ambiguity of some sentences which contain embedded

clauses can hardly be accounted for in the constituent structure

model. To this end consider the following sentences:/ X(4) Nguruve iyi yakurisa kudva.

This sentence may mean either that the pig is so old that it

cannot eat or that the pig is.so old that its meat is difficultto be eaten* That is, on one reading the noun phrase n&uruv© iyl

/is the subject of the verb kudya while on the other reading it

is a direct object. The analysis of sentence (4 ) would be something like this:

/ ■ t(5) Nguruve iyi yakurisa kudya.

1 1 I 1

Whatever the correct analysis of this sentence may be according to the structuralists, the crucial point nevertheless is that there is no way of showing in the analysis itself that the sentence has

two semantic readings. Yet it is essential to describe ho?/ this sentence came to be ambiguous.

10

A similar problem arises in analysing sentences which are

structurally homonymous like those in (6) and (7)./ / / / /(6) Hulcu yacho inoshupa kudya.

('The chicken is difficult to eat)/ n / / / / / fv7) Hulcu yacho inoda kudya.

(The chicken likes to feed.)

A constituent type of grammar would analyse these two sentences in the same way that (5) is analysed. That is, (6) and (7 ) wouldhave an identical analysis. But our intuition as native speakers

f y / /of Shona reveals that in (6) the noun phrase huku .yacho is a direct/

object of the verb kudya, while in (7 ) the same noup. phrase is a /subject of kudya. Therefore if our grammar is to be descriptively

adequate:it must be feflected in the analysis that the grammatical/ / / /relationship between huku yacho and kudya is different in the

two sentences (6) and (7)«In any given sentence we would like our grammar to be

able to describe the relationship that obtains between the various

constituents, How if we examine sentences (8) and (9 ) below we/ / / /

notice that the strings achifara and yakatengw^a lcwa Clot ora actually/ / 7belong to the subject noun phrases murume and mombe iye respectively

to which they refer./ / / / y / / / /(8) Murume akaita basa rake achifara.

(The man did his work happily.). / / / / / / /(9) Mombe iye yakarumwa nebere, yakatengwa lcwaGotora.

(That cow was killed by a hyena, which was bought

from Gotora's kraal.)In a constituent type of grammar it would be difficult to show that

/ / / / ’the complement clauses achifara and yakatengwa kwaGotora belong tothe subject noun phrases of these sentences, or at least that they

start off from these subject noun phrases.

11

Another process which' a constituent structure type of

grammar seems ill-equiped to handle is the transformation

commonly referred to in the literature as raising. Towards

this end consider £he sentences in (10),, N / / / / / / //(10) a. Yanhu vanoziva kuti ndiwe wakaita zvinhu izvi.

(The people know that it was you who; did these things.)V / / / / / / / //b, vanhu vanokuziva kuti ndiwe wakaita zvinhu izvi.

(The people know you to have done these things.)

T h e : two sentences in (lO) mean more or less the same thing. The/ / / / //direct object of the verb -ziva in (10a) is kuti ndiwe wakaita

. / t \zvinhu izvi and the noun phrase iwe (you ) is part of the embeddedt ■ \ / / *clause. But in (10b) the direct object of -ziva is iwe which is

represented by its pronominal form -ku- . I do not know how a

constituent type of grammar would cope with' the analysis of

these two sentences while at the same time accounting for

their synonymity.

0.2. In part this study is a test of the theories, as

applied to Shona, that have been propounded in the generative

transformational grammar model concerning sentential complements.

The study, though basically carried within the framework of

Aspects of the Theory of Syntax, incorporates much of the

modification which that work has undergone since its publication

in 1965*

5 . N. Chomsky in Aspects of the Theory of Syntax, M.I.T. Press,

Cambridge, (1965)-

121 .0 SENTENTIAL COMPLEMENTS OF VERBS

1.1 It is a generally accepted fact that the number of sentencesin any, language is potentially infinite. That this is true is

easy to comprehend because sentences may be constructed withother sentences embedded, in them. This process of recursionis one of the properties of language. Through recursion the

1phrase structure rules which generated the first sentence may be applied again to produce an embedded sentence, and the process can be repeated several times. The most commonly discussed

processes of recursion are conjunction, relativization and sentential complementation (or predicate complementation, as it is

sometimes called). This study will be concerned almost exclusively with sentential complementation. Sentential complementation may be divided into noun phrase complementation and verb phrase complementation,1 .2 The Case Eor Noun Phrase Complementation

In order to appreciate more fully the arguments which follow not only in this chapter, but in later chapters as well, it may be as well to discuss first how we can discover (a) noun

phrases which are in a subject position and (b) those which are in an object position.1.2.1 Tests for noun phrases in subject position

Although there may be many tests for discovering noun phrases in subject position only two will be discussed here.

The quantity of tests by itself is inconsequential. What is important is to be able to recognize a noun phrase that, is

functioning as a subject .

1. See Appendix A for the phrase structure rules of Shona.

131,2,1.1 First, if there is concordial agreement between a nounphrase and a predicate, that noun phrase is functioning as a subject

In other words, if a noun phrase has its class feature copied on to the auxiliary in the structural change of the Gender Copying rule, that noun phrase is in a subject position. For instance, giventhe following sentence:

/ \ J i f f / i(1) Vakomana vanotamba nhabvu.

(Boys play football,)

it may be represented roughly by the following tree in (2), There is an obligatory rule in Shona which states that whenever there

is a noun phrase which occurs before AUX, copy the class feature of that noun phrase on to AUX as in (j) below. The rule of Gender Copying will be presented formally in Chapter 5®

(2)

Tense

tambakomana nhabvu

(3)

Tense

nhabvuI-—

komana no+N -fCL+C

tamba

Here are more examples to illustrate what is meant by concordial agreement between a noun phrase in subject position and its predicate/ \ • / / ?(4 ) a® Chrngwa (7 ) chakaora .

(The bread is mouldy*)

2, The number placed after a noun phrase here and in the rest of this study indicates the noun class to which the noun phrase belongs.

14/ * fb*. Zvengo (8 ) zvawa>

(The walls fell down#)

c. Masai (6 ) achapuWz ik al(The eggs will get broken..)

/ , . / / / / , / / /d. Rufn (11.) runouya kumunhu wese wese..(Death comes to every person.)

a*. Mumfha' (18) maitonhora.(inside the house - it was cold = it was, cold inside the house.)

t1*2.1*2 The second test involves the interrogative pro-forms ani and✓ /chii* The form ani is used for human noun phrases while chxi is used

for non-human noun phrases. In this case the interrogative pro-forms,in. addition to substituting for a noun phrase, are stabilised, or inother words turned into a predicate, with the rest of the sentence

/becoming a relative clause. With regard to ani this is done by the useof the formative ndi. If, for instance, we wish to find out whether ornot vakomana in (1 ) is a subject noun, phrase, we first of all find outwhether it has the feature ^human/ or not. In this case we see that it

has the feature /human/.. As we saw earlier on the interrogative pro—/form which we need in this case is ani. Applying what has been said

above, (1.) is transformed into (5 ).✓ x / / / / / / , / A(5) Ndiani anotamba nhabvu? (reply: vakomana)

(it is who - who plays - football =Who plays football? (reply: boys))

tNote that ani as such has singular reference and this explains why we have the. class feature a rather than va before the tense sign no in

(5). Notice also the difference in their tone pattern between vanotamba/ / / ✓ inhabvu in (1) and ano tamba nhabvu in O ) . The former is In the indica­

tive mood while the latter Is in the relative mood. However this is

not to say that this is the only difference

15there is between the indicative and relative constructions.

If we want to find out whether chingwa in (4a) is a subject noun phrase or not we proceed in the same way. Here we discover that chin&wa has the feature ^human/, Therefore

/the interrogative pro-form which we are going to employ is chii.We transform the structure which underlies (4a) into that whichunderlies (6).

f / /(6) Chii chakaora? (reply: chingwa)(lit. it is what - that is mouldy=

What is mouldy?) (reply: bread)In this case the result of the stabilization or predication of

/chii consists of raising the tone on the class affix chi and lowering the tone on the following syllable. Here are further

illustrative examples;, \ / v / , , J / /(7) a. (i) Mwana (1) akarara.

(The child is asleep.)/ N / J / / A(ii) Ndiani akarara? (reply: mwana)

(lit. it is who - who is asleep =Who is asleep?) (reply: the child)

/ \ / f \ n / f / /b. (i) Sara (1a) aenda kunochera mvura.

(Sara - went - to' fetoh - water =Sara has gone to fetch some water.)

. . ' / / / / / / / / . /.(ii) Hdiani aenda kunochera mvura? (reply: Sara)(it is who - who went - to fetch - water =Who has gone to fetch some water?) (reply: Sara)

c. (i) Harahwa (9) yaitonhorwa nechando.(old man - was being cooled - by cold =

The old man was feeling cold.)

, , * * / / / /(ii) Ndiarii aitonhorwa nechando?

(it was who - who was being cooled

Who was feeling cold?), V t . v / / /(i) Mukuyu (5) wakapunzika.

(fig-tree - fell down =

The fig-tree fell down.)f /(ii) Chii chakapunzika?

(it is what - that fell down =

What fell down,)/ a / / \ / / /(i) Mango (9) yakaibva.

(mango - is ripe

The mango is ripe.)

(ii) Chii chakaibva?

16(reply: harahwa)

by cold =(reply: the old man)

(reply: mukpyu)

(reply: the fig-tree)

(reply: mango)(it is what - that is ripe =

What is ripe?) (reply: the mango)1.2.2 Tests for noun phrases in object position

For discovering noun phrases which are in object position four procedural methods will be discussed in this subsection.

I.2.2.1 The Passive testLet us consider the sentence below:

f / I f(8) Toni anoda Koni.(Toni loves Koni.)

/ /In this sentence there are two noun phrases, namely Toni and Koni.If this sentence is transformed into the passive it becomes

. , ( i f / /(9) Koni anodiwa naToni,

(Koni « is loved - by Toni =

Koni is loved by Toni.)What the passive transformation did was to interchange the noun

/ / /phrase Toni and Koni« Koni is now the grammatical subject. The

claim made by passive test is that the string which is made the

17

subject by transformation is an object noun phrase. Nowconsider the following sentences:, . / / / /(10) a. Mwana akuvara ruoko.

(child - got hurt - arm ='The child hurt its arm.)

/ / / / / xb. *Ruoko rwakuvarwa nemwana.'(arm - was hurt - by child*Its arm was hurt by the child.)

/ \ / / / / / /(11) a. Bonda akacheka vana nzara.(Bonda - cut - children - finger-nails =Bonda cut the children's finger-nails.)

f t / / / ' rb. Yana valcachekwa nzara naBonda.(children - were cut - finger-nails- by Bonda =

The children had their finger-nails cut by Bonda.)/ / / f / {c. *Nzara dzakachekwa vana naBonda.

(finger-nails - were cut - children - by Bonda -

*Finger-nails were cut the children by Bonda.)/ t / / / /(12) a. Ambuya vakapa muzukuru doro k a v i n ,

(grandmother - gave - grandson - beer - twiceThe grandmother .gave her grandson some beer'twiee.) /. .

( J / / / / /b, Muzukuru alcapihwa doro kaviri naambuya.(grandson - was given -beer - tv/ice - by grandmotherThe grandson was given some beer twice by his grandmother)

/ / / / / / /c. Doro ralcapihwa muzukuru kaviri naambuya.(beer - was given - grandson - twice - by grandmother = Beer was offered to the grandson twice by his grandmother)

3. The asterisk (*) will be used throughout this work to signify

an ungrammatical sentence.

18i f f / / / /d. ^Kaviri kakapihwa muzukuru doro naambuya.

(*twice - was given - grandson - beer - by grandmother.)rt / / / \ f ' *^y the passive transformation ruoko in (10b), vana and nzara in

(11b ) and (11c), and muzukuru, doro and kaviri in (12b), (12c)and (l2d) respectively have become the new grammatical subjects.The ungrammaticality of (10b) indicates that ruoko in (10a) is

/ \ /not an object noun phrase. In (11a) only vana is an object noun

phrase as is evidenced by the grammaticality of (11b), and the

ungrammaticality of (11c). The v/ell-formedness of the passivesentences in (12b) and (12c) shows that muzukuru and doro are

/object noun phrases. Kaviri, although it is a noun phrase of

some kind, at least in form, is not an object noun phrase as the ungrammaticality of (l2d) shows.

Locative noun phrases occurring after the verb can be transposed to the subject position by the passive transformation indicating that according to this test such locative noun phrasesare to be regarded as object noun phrases.

/ J s I / / /(13) a. Jeke akaona nyoka kurukova.(Jack - saw - a snake - at river

Jack saw a snake at the river.)i* / / / / / / /b. Nyoka yakaonekwa kurukova naJeke.

(snake - was seen - at river - by Jack

A snake was seen at the river by Jack.)t f f f f / /c. Kurukova kwakaonekwa nyoka naJeke.

(at fiver - was seen - a snake ~ by Jack =A snake was seen at the river by Jack.)

, , f 1 / / / / / /(14) a® VekwaChari vanochera mvura patsime iri.

(Chari’s people ~ fetch - water - at well - this Ghari1 s people get their water supply from this wrell. )

/ / t / / / / / /Mvura inocherwa patsime iri nevekwaChari.

(water - is fetched - at well - this - by Chari's peopl

Water is fetched from this well by Chari'people.)/ / / / / / / / /Patsime iri panocherwa mvura nevekwaChari,

(at well - this-is fetched - water - by Chari's people

Chari's people fetch their water from this well.)/ / / / / / /Sekai alcakanda Toko bhora kuchikoro.

( Sekai - threw - Toko - ball - at school =Sekai threw a ball to Toko at school.)

/ t / / / / /Toko akakandwa bhora kuchikoro naSekai. .(Tolco - was throv/n - ball - at school - by Sekai =Toko was thrown a ball at school by Sekai.)

/ / / / / / / /Bhora rakakandwa Toko kuchikoro naSekai.

(ball - was throv/n - Toko *- at school - by Sekai =ball was throv/n to Toko at school by Sekai.)

/ / / / / / /Kuchikoro kwakakandwa Toko bhora naSekai.

(at school - was thrown - Toko ~ ball - by Sekai =Sekai threw a ball to Toko at school.)

‘ / / / /' / /Zanza akaudza Mafaro nyaya rye kubhawa.(Zanza - told - Mafaro - story - that - at beer hall =

Zanza told Mafaro that story at the beer hall.)/ / / / i t /Mafaro akaudzwa nyaya iye kubhawa naZanza.

(Mafaro - was told - story - that - at beer hall -

by Zanza =Mafaro was told that story at the beer hall by Zanza.)

/ / / / / • / /Nyaya iye yakaudzwa Mafaro kubhawa naZanza.(story - that - was told - Mafaro - at beer hall -

by Zanza.'- hat story was related to Mafaro at the beer hall

by Zanza.)

ao

/ / / / / / /d. Kubhawa kwakaudzYfa Mafaro nyaya iye naZanza.(at feeer hall - was related - story - that -Mafaro -

hy ZanzaMafaro had that story narrated to him at the beer hall by Zanza.)

Although a locative noun phrase which occurs after the verb canbe transposed to the subject position by the passive transformation,there seem to be degrees of acceptability of the resulting sentencesdepending on the number of other noun phrases occurring beforeit in the input sentence. For instance, in (13a) and (14a) the

/locative noun phrases kurukova and patsime iri respectively eachconstitutes the second object noun phrase. But in (15a) and (16a)

/ /the locative noun phrases kuchikoro and kubhawa respectively each

forms the third object noun phrase. The locatives which are mentioned above form the grammatical subjects of the passive

sentences in (13c), (14c), (l5d) and (l6d), respectively. Nevertheless,

we know intuitively that the sentences in (13c) and (14c) are more acceptable than those in (13d) and (l6d). Still all the four sentences are grammatical. However since locative complements can occur after almost any verb in Shona one wonders whether they ought really to be looked upon as object noun phrases.

A noun phrase which expresses time does not appear to be confirmed as an object noun phrase by the passive test as the

following examples will illustrate ;, / / / / / ( /(17) a. Mukomana akabata mbeva nezuro.

(boy - caught - rniob - yesterday =The boy caught some mice yesterday.)

21

/ / / f / / / /b, *Nezuro alcabatwa mbeva nemukomana.(*Yesterday was caught some mice by the boy.)

/ s / / J / / / / /(18; a. Mapurisa achasvika kumba kwenyu mangwana.

(poli cemen - will arrive- at home - your - tomorrow =The policemen Vvrill get to your house tomorrow.)

/ / / / / / / / ✓’b. Kumba kwenyu kuchasvikwa mangwana nemapurisa.(at your home - v/ill be arrived - tomorrow - by policemenThe policemen will get to your house tomorrow.)

/ / / / / / f / /c. *Mangwana achasvikwra kumba kwenyu nemapurisa.(* Tomorrow - will be- arrived - at your house - by the

■: policemen.)

I,2.2.2 The interrogative Pro-form testThis test involves substituting the interrogative / / fpro-forms ani? and chii?lfor the noun phrase which we suspect to

be an object noun phrase. If the answer to the resulting question is a grammatical sentence, that noun phrase is an object noun phrase,

The examples in (19) - (22) show how this test works./ \ f / / / , / N v(19) a. Toni anoda Koni. (see (8))

/ j / /b. Toni anoda ani*j! (reply: Koni)(Toni loves whom?) (reply; Koni)

. / / / / / / / , / \ \(20) a. Mwana akakuvara ruoko. (see (10))

b. Mwana akakuvara chii? (reply; ruoko)(child - was injured-what =

The child was injured where?) (reply: on the arm)/ \ / / / / t / / t w(21) a. Bonda akacheka vana nzara. (see (11))

f J / / J / f Ab. Bonda akacheka ani nzara? (reply: vana)

(Bonda - cut - whom - finger-nails -Bonda cut whose finger-nails? (reply: the children's)

/ / / / / / Ac. Bonda akacheka vana chii? (reply: nzara)

22

(Bonda - cut - children - what =Bonda cut the children’s what? (reply: finger-nails)

(22) a*. Ambuya vakapa muzukuru doro kaviri^ (see 1.2)b. Ambuya vakapa ani'' doro kaviri? (reply: muzukuru)

(grandmother - gave - whom - beer - twice =

Who did grandmother give beer twice? (reply: her grandson)t i / / / / / . .c. Ambuya vakapa muzukuru chii kaviri? (reply: doro)

(grandmother - gave - grandson - what - twice =What did grandmother give to her grandson twice?

(reply: beer))

d„. * Ambuya vakapa muzukuru doro chii? (reply: kaviri)(*The grandmother - gave - her grandson - beer - what?)

(reply: tv/ic;e))/ / /Observe that this test will confirm ruoko and nzara in (20) and (21)

respectively as object noun phrases* But in actual fact these cannott fbe said to be object noun phrases in the true sense* To say that ruoko

here is an object noun phrase is counter-intuitive. In this respect it would appear that this test is less reliable than the passive test. Nonetheless in view of the fact that the nature of the complements of the verbs of the type exemplified by -kuvara in. .(2 0) and -cheka in (21)^ is not clearly understood at this stage in Shona linguistic studies one cannot make a categorical statement as to which of these two tests is more reliable than the other.

Locative noun phrases occurring after the verb will be rejec­

ted as object noun phrases by the interrogative pro-form test./ / / / / ✓ /(2 3) a.. Jeke akaona nyoka kurukova. (see 1 3)

Af. for a commient on verb phrases featuring verbs of this type see

p- 180 - 182.

23/ / / / / f /b. *Jeke akaona nyoka chii? (reply: kurukova)

(*Jack - saw - a snake - what)/ / / / / / / /(2 4) a. VekwaChari vanochera mvura patsime iri. (see (14) )/ / / / / / / ' /b. *VekwaChari vanobhera mvura chii? (reply: patsime iri)

(*Chari's people - fetch - water - what?)/ / / / f t /(2ft) a. Sekai akalcanda Toko bhora kuchikoro* (see(l5) )/ / / / / / / /

b. *Se}jai akalcanda Toko bhora chii? (reply: kuchikoro)

(*Sekai - threw - Toko - a ball - what?)As was the case faith $he passive test noun phrases which

indicate time are not confirmed as object noun phrases by this test., * / / / / / / f . t \ \(26) a. Mukomana akabata mbeva nezuro. (see (17) )

/ 1 * / / / ,b. *Mukomana akabata mbeva chn? (reply: nezuro)

(*The boy - caught - mice - what?)(27) a* Mapuiisa achasvika kumba kwenyu mangwaha (see (18) )

/ / j / / / / / / .b. *Mapurisa achasvika kumba kwenyu chii? (reply: mangwana)

(--Policemen - will arrive - at home - your - what?)

I.2.2.5 The Object Anaphora testThe transformational rule of Object Anaphora is discussed

in some detail in (5-13)* What it is briefly is that in Shonasome noun phrase complements of verbs may be represented by their

pro-forms (or object prefixes as they are often referred to inBantu linguistic studies) if the proper structural descriptionfor this rule is met. Such pro-forms are normally reflexes of

the noun prefix. The claim that is made by this test is that anoun phrase complement which can be represented by its pro-formby this transformation is an object noun phrase. The application of

this test to the sentences (8), (10), (ii), (12), (13) ? (15) and(17) yields the sentences in (28) - (34) respectively.

(28)

(29)

(JO) a

b

(31) a

(32) a

b

(33) a

To

Zk

/ t / / iToni anomud a . (referring to K o n i )

(Toni loves h e r .)/ / / / / / , /*Mwana akarukuvara. (referring to r u o k o )

(*The child hurt i t )/ ' / t / / . ABonda akavacheka nzara. (referring to vana )

(Bonda - cut them - finger-nails =

Bonda cut their finger-nails.)f / / / / / t /hBonda akadzicheka vana. (referring to nzara )

(Bonda - cut them - children -

*Bonda cut them the children.)/' / / / / t /

Amhuya vakamupa doro kaviri. (referring to muzukuru)

(grandmother ~ gave hijii - beer - twice The grandmother gave him beer tv/ice.)

/ / * / / / / , VAmbuya vakaripa muzukuru kaviri. (referring to doro)

(grandmother - gave ijt - grandson - twice =

The grandmother gave ijt to her grandson twice.)/ ' / / f / . \*Ambuya vakakapa muzukuru doro. (referring to kaviri )

(*The grandmother ~ gave ijt - her grandson - some beer.)/ / '/ / / , /Jeke akaiona kurukova. (referring to nyoka )

(Jack - saw _it - at river =

Jack saw _it at the river.)/ / / / ' ✓ / , .Jeke akakuona nyoka. (referring to kurukova)

(Jack - saw at ij. - a snake =

Jack saw a snake there.). / / / / / / , KSekai akamukanda bhora kuchikoro, (referring to T o k o )

(Sekai - threw her - a ball - at school

Sekai threw the ball at her at school.)/ ' ' / / / / , / VSekai akarikanda Toko kuchikoro; (referring to b h o r a )

(Sekai - threw it - Toko - at school =

Sekai threw it. to Toko at school.)

25/ *' * / / /c. /Sekai akakukanda Toko bhora. (referring to kuchikoro)

(*Sekai - threw at it. - Toko - a ball =

*Selcai threw the ball to Toko at school*)/ \ ! i f * t t f f(34) a* Mukomana akaibata nezuro. (referring to mbeva)

(boy - caught it. - yesterdayThe boy caught it. yesterday.)

/ / / / / / / /b. *Mukomana alcamubata mbeva. (referring to nezuro)(*The boy - caught i_b - the mouse.)

Notice that according to this test, as according to the passive

testjboth ruoko/ and nzara'' in (2 9) and (30) are not object noun phrases. Sentence (33<?) is very low on the scale of acceptability. Most speakers would reject it as ill-formed, nevertheless it is

still grammatical. This seems to confirm the remark which was made in (1.2.2.1) above concerning locatives. The passive test and

the object anaphora test seem to confirm object noun phrases of the same type*I.2.2 .4 The Pseudo-cleft test

This test operates roughly as follows: the noun phrase v/hich is suspected to be an object noun phrase is taken out from

its position and placed at the beginning of the sentence. The new

position it now assumes in the sentence may be immediately before or after the subject noun phrase. This is followed by the stabilised form of its pro-form , which in turn is followed by the rest of the sentence in the objective relative. This is shown in the

examples below. The noun phrases which are being tested are

underlingd./ / i f(35) a* Yana vakaba mango.

(The children stole some man^o.aa.)/ t / / /b. Yana, mango ndidzo dzavakaba.

(children - mango - it was they - which they stole =

It was some mangoes which the children stole.)

26, ., / t / t / ' / / / /(36; a. Amai vakaratidza mwanasikana wavo zvinhu zvekuita.

(mother - showed - daughter - her - things - of doing

The mother showed her daughter what to do). / / f / f / / 'b, Amal, mwanasikana wavo ndiye wavakaratidza

/ zvinhu zvekuita*

(Mother - daughter - her - it was she - whom she

showed - things - of doing

It was her daughter whom the mother showed what to do.)/ / / / / /c. Amai, zvinhu zvekuita ndizvo zvavakaratidza

7 -i 7 /mwanasikana wavo.(mother - things of doing - it. was they - which she showed - daughter - her It was what to do that the mother showed her daughter.)

However this test by itself Is unreliable because it will confirm as object noun phrases some noun phrases which will not be passed

by the other tests as the following examples show:

(37) Mwana/ akuvara ruoko/. (see (10a))/ / / /' / / /b. Mwana, ruoko ndirwo rwaakuvara.

(child - arm - it is it - which she hurt

It was the ai?m which the child hurt 'itself.). / / / / / / , t w(38) a. Honda akacheka vana nzara. (see (11a;;

/ / / / / / /b. Honda, vana ndivo vaakacheka nzara.(Bonda - children - it was they - whom he cut - finger”

nails =It was the children whom Bonda cut their finger-nails.)

/ / / f t /c. Bonda, nzara ndidzo dzaakacheka vana.(Bonda - finger-nails - it was they - whom he cut -

childrenIt was the children’s finger-nails that Bonda cut.)

This is however more of a test for discovering noun phrases in

2?

a sentence than for discovering object noun phrases. It has

been included here though as it will help establish that some

sentential complements are actually under the direct domination of NP.

Of these tests the passive and the object anaphora appear to be the most reliable, with the pseudo-cleft as the least

reliable, llaybe further investigation Into the way object noun phrases behave in sentences will result in the discovery of a more ..'rigorous test or tests. For the purpose of this study

however identification of object noun phrases will be based on these four tests.1.2.3 Noun Phrase Complements

Turning now to our discussion of noun phrase complement let us consider the following sentences:

, N f f / / /(39) Mombe dzako dzadya chibage changu.(cows - your - ate - maize crop - my =Your cows ate up my maize crop.)

/ / / / / /b. Chibage changu chadyiwa nemombe dzako,(maize crop - my - was eaten - by cows - your =

My maize crop was eaten up by your cows. )It is uncontroversial that these two sentences are related toeach other structurally. One Is an active and the other apassive sentence. Semantically they are synonymous. Now

consider the following sentences:/ ' f / / / / / /(40) a. Mwoyo anoziva kuti Kondo akaba mari.

(Mwoyo - knows - that - Kondo - stole - money =Mwoyo knows that Kondo stole some money.)

/ <' s / ( / / / /^ * Kuti Kondo akaba mari zvinozivikanwa naMwoyo,

(that - Kondo - stole - money - is known - by Mwoyo That Kondo stole some money is known by Mwoyo.)

28

The two sentences In (40) are comparable to the two sentencesin (59)* That is, one is active and the other is passive. Butwhereas (39a) has as its direct object a noun phrase, (40a)

/ / / Ihas a sentence:, namely, kuti Kondo akaba mari, as its verb complement. For convenience such sentences as the one underlined

which are embedded in others will be referred to as complement sentences or complement clauses. The word kuti in this complement sentence cannot he regarded either as a noun or as noun phrase in

its own right because It-cannot occur by itself in either a subject or an object position as noun phrases normally do. Furthermore as we shall see later the word kuti has not thefeature /i-class/^ which every noun, in Shona has. So what we have

in (40a) is a verb which is accompanied by a sentence (s) as direct object. This being the case the passive rule, for instance, has to be altered'in order to accommodate this S object. That

this object is an S of some sort can be shown by, say, applying

the passive rule within it. If this is do^e the result is sentence

(41) -. / / / / / / / / / /(41) Mwoyo anoziva kuti mari yakabiwa naKondo.

(Mwoyo - knows- that - money - was stolen - by Kondo =Mwoyo knows that the money was stolen by Kondo.)

This is by no means the only way of showing that the string / / / / fkuti Kondo akaba mari is an S of some kind. This string can

undergo the transformational rules of pseudo-cleaving, interrogative

pro-form and object anaphora as is shown in (4 2 ), (4 3 ) and (4 4)respectively. All these transformations apply within sentences.

/ / / / / / / / /(4 2 ) Mwoyo anoziva kuti Kondo ndiye akaba mari.

(Mwoyo - knows - that - Kondo it is he - who stole -

money =

Mwoyo know^ that it was Kondo who stole the money.)

29/ \ / / / / / / ■ « ' ' / f f * x(43) kwoyo anoziva kuti Kondo akaba chii? (reply: mari)

(Mwoyo - knows - that - Kondo - stole - what =Mwoyo knows that Kondo stole what? (reply: the money) )

/ \ l * f i / / / / / , \(44) Mwoyo anoziva kuti Kondo akaiba. (referring to mari)(Mwoyo - knows - that - Kondo - stole it =

w .Mwoyo knot's that kondo stole it. )Let us assume that the passive rule in Shona operates as follows

(45) X - Passive - KP1 - AUX - V - Y - NPn - ZS D r 1 2 3 4 5 6 7 8

„ >SC: 1 0 7 4 5+W 6 0 8 na + 3

Now if the direct object of a verb can be an S as well as an NP there is need to modify our formulation of the passive rule inorder to accommodate cases of an object sentence- This is done

in (4 6) tel ow:

(4 6 ) X - Passive - NP, - AUX - V - Y - (NP> \- Z

SD:

SC:

But now examine the following two sentences:, . / * / / / / / / / / / /(47) a. Kuti asvike kuno zvinoratidza kuti arioda mukadzi wake,

(that - he arrive - here - shows - that - he loves -

wife - hisFor him to come here shows that he loves his wife.)

s / / / / / / . / .b. Kuti anoda mukadzi wake zvinoratidzwa nekuti/ & / / / asvike kuno.

(that - he loves - wife - his - is shown - by that -

he arrives - here =That he loves his wife is shown by his coming here.)

30

/ % / / /The string kuti asvike kuno is the subject of the sentence

in (47a)* But this subject consists of an S. Sentence (47b) is the passive sentence of (47a). So if S can also function as a subject, this ought to be reflected in our formulation of the passive rule which is modified below as (48)*

(48) X - Passive -|NP..V AUX - V - Y ~CNP2")- Z

( s / [ s

SD: 1 2 3 4 5 6 7

SC: 1 0 7 4 5+w 6 0 8 na + 3 P

At this stage, although our grammar may be adequate from an observational viewpoint, it reveals that generality is very

much diminished. This is brought about, not because a string/ ^ i t tlike kuti asvike kuno is an S of some sort, but because of

the assumption that such a phrase cannot be regarded as a noun

phrase. If in our grammar UP is permitted to dominate S as well as N (noun), then the original formulation of the passive rule in (4 5) will effectively generate the sentences in (39b), (40b)

and (47b)«Now let us see what phrase structure rules are

required for noun phrase complementation. As for any sentence

the following rules apply:

(4 9) 3 * S ---- — — > NP + AUX + PDP

b . PDF -— 5> VP (PP)(APV)

c . VP - Y (NP)

Then in order to produce the structures which underlie complement

sentences there is need for the following additional rule:g . S fP N ( S ) •

3b

'The expansion of NP in (49d) differs from another possible expansion of N P , namely, the one which produces relative clauses

(up ------------ y NP + S)5 .

There are however instances in the language where phrases which look like those which have just been discussed (i.e. they are to be analysed as instances of S at some level) do not behave like noun phrases. Such are the complement sentences in (5 0) - (5 2)./ \ i t / / /sy. / /(50) a. Sekuru ava kuti aende nhasi.

(uncle - is now - to say - he goes - today =Uncle now says that he should go today.)

/ f / / / /b. *Kuti aende nhasi zvaviwa nasekuru.

(*To say - he goes - today - .is now being ~ by uncle)t f / / f / fc / / .c. *Sekuru ava chii? (reply: kuti aende nhasi)

(*Uncle - is now - what?)/ / / , / / /d. *Sekuru akuva. (referring to kuti aende nhasi)

(*Uncle - is now it.)/ \ * t / / /(51) a. Ali akanyatsa kurova Foreman.

(Ali - did well - to beat - Foreman =Ali clearly defeated Foreman.)

/ / f t /b. *Kurova Foreman kwakanyatswa naAli.(*To beat - Foreman - was done well - by Ali.)

/ / / / / / / Nc. *Aii akanyatsa chii? Ireply: kurova Foreman)

(*Ali - did well what?)

5. For a discussion of the vsri.os analyses of the relativeA

construction see Major Syntactic Structures of English,

R.P. Siockwell, Paul Schachter and B. Hall Partee (eds).

32/ / / / f t \d. *Ali akazvinyatsa. (referring to kurova Foreman)

(*Ali did it well.)f v / / ( / ( { t ( t /(52) a, Garwe anofanira kuti auye mangwana.

(Garwe - ought - that - he comes - tomorrow =Garwe must come tomorrow.)

/ / f / f / / /h. fKuti auye mangwana zvinofanirwa naGarwe.

(*That - he comes - tomorrow - is ought - by Garwe.)/ / / ■ / / f . f / / /'*Garwe anofanira chii? (reply: kuti auye mangwana)

(*Garwe - ought - what?)f J • f f f t / / / /'d. *Garwe anozvifanira. (referring to kuti auve mangwana)

(*Garwe - ought it.)

In the sentences (50)> (51) and (52) attention was centred on the/ i f f / / / / /complement sentences kuti aende nhasi, kurova Foreman and kuti auye

mangwana respectively. The (b), (c) and (d) sentences in (50) - (52)

show the results of the passive, the interrogative pro-form and the object anaphora transformations when applied to the deep

structures which underlie the (a) sentences. The sentences in

(b) - (d) are all ill-formed.If these transformations are applied to the structure

which underlies (4 0) the sentences which result are those in (5 5)

below./ * f t / f / / / , . . .(5 3) a. Mwoyo anoziva kuti Kondo akaba m a n . (see (40a))/ / / / / / / / / b. Kuti Kondo akaba mari zvinozivikanwa naMwoyo.

(That Kondo stole some money is known by Mwoyo.)/ f /

Mwoyo anoziva chii

(Mwoyo knows whatt)t f

Mwoyo anozviziva

(Mwoyo knows i t .)

i f f / , / ( ( f tMwoyo anoziva chii? (reply: kuti Kondo akaba man )

( / 1 L / fMwoyo anozviziva. (referring to kuti Kondo akaba m a r i ).

3/ / / / / f /Although the strings kuti aende nhasi and kuti auye mangwana

in (50) and (5 2) respectively are similar to the string kuti kondo / / / . Nakaba mari in (55) in their structure, they nevertheless differfrom it in that they do not behave like noun phrases as it does.

/ / / / /It is observed that the string kuti Hondo akaba mari passes

these three tests for object noun phrases as is shown in (55)*So this string is an instance of an object noun phrase. That

t t / / /the string kuti Kondo akaba mari is functioning as a noun phrase, in this sentence is further substantiated by the pseudo-cleft

transformation as indicated in (54) •/ v / / / / J f / /(54) Kuti Kondo akaba mari ndizvo gvaanoziva Mwoyo.

(that - Kondo - stole - money - it is it - what he knows -

Mwoyo -That Kondo stole some money Is what Mwoyo knows.)

Notice that the three complement sentences from (5 0) to (52)

inclusive all fall the tests for object noun phrases.If, for argument's sake, It is still maintained that NP

may not dominate then It means that, for instance, in ourpassive rule in (48), we have to add a caveat which states in a rather uijclear manner that only sentences such as are found in (40a)

and (47a), and not those in (5 0), (51 ) and (5 2), may feature in a passive transformation.

However, if on the other hand we allow NP to dominate

S as well as N, then the complement sentences in (40a) and (47a)

will be regarded as instances of S which is dominated by NP, while

those in (50), (51) and (52) will be instances of S under the direct domination of VP. That is, the complement sentences in (50) - (52) will not be dominated by NP, hence the passive «,(( vAe,

O'&e-r- transformations which are employed as tests for object noun phrases

will block. That being the case, there will be no « ,c no need to e n t ^

3k

S separately from NP in the passive rule as was done in (48).The original formulation of the passive rule in (45) willsuffice. There will he no need to differentiate in the phrase

/ /structure rules which introduce the string chibage changuin (59®; and kuti Kondo akaba m a n m (40a; in the derivationof these two sentences. (Of course once both have beenintroduced as NP by the phrase structure rules, the two noun

phrases may now be expanded differently.) Indeed to make thisdistinction would be an admission of failure to capture alinguistic generalization, and that would weaken the grammar

considerably. On the other hand, not to recognize that the/ / / / ^string Kuti Kondo akaba mari is an instance of S at some stage

In the derivational history of the noun phrase is equally a mistaken approach. If the approach suggested in this paragraph is adopted, the caveat which was mentioned above falls away

as it becomes unnecessary. Furthermore, if NP dominates S in

(4 0), but not in (5 0), (5 1) and (5 2), then our grammar automatically predicts that (40b) is grammatical, while (50b-d), (51b-d) and

(52b-d) are not.The approach which says that an NP can dominate an S

will help to explain at least one phenomenon in Shona, namely, the behaviour of the kuti clauses. To this end consider the

following sentences:/ \ / f ' /(55) a* Kuti udzokere lcumusha zvinonetsa.

(that- you return - home - is worrying =For you to go back home is difficult.)

/ / / /b. Zvinonetsa kuti udzokere kumusha.

(it is difficult for you to go back home.)

35

f s \ t f i i / / / / t ,(.5°/ a- Mwoyo anoziva kuti kondo akaba mari» (see (40a);/ / / ! ( f t * fb» Kuti Kondo akaba mari zvinozivikanwa naMwoyo. (see (40b))

/ / / f / ( ( t /c.. Zvinozivikanwa naMwoyo kuti Kondo akaba mari.(it is known by Mwoyo that Kondo stole some money.)

The sentences in (55) are structurally related to each other,

and so are those in (56)» The kuti clauses in both (55) and(5 6) are instances of S under the domination of NP. It is aninteresting observation that a kuti clause in a subject position

may optionally be postposed to the end of the sentence. In (55b) it is a kuti clause in a subject position in an active

sentence that has been postposed, and in (56c) it is a kuticlause in a subject position but in a passive sentence that has been so postposed. This postpositioning of a subject kuti

clause, one might say, is not different from the postpositioning of an ordinary NP, for in Shona noun phrases in subject position

are not infrequently placed at the end of their sentence as

the following sentences show.* / / / /(57) a. Mombe dziye dzaneta zvino.

(oxen ~ those - are tired - now =Those oxen are tired now.)

/ / / / / /b. Dzaneta zvino mombe dziye.( they are tired - now - oxen - those =Those oxen are now tired.)

. / / / / /(5 8) a. Vana vadilci havaende kubira.(children - little - do not go - to biraLittle children do not go to the bira.)

/ / t / /b. Havaende kubira vana vadiki.(they do not go - to the bira - children - little

Little children do not go to the bira.)

36There seems, nevertheless, to be a difference between the kuti clauses

/ / fand the other noun phrases like mombe dziye in (37) vana vadrki in(38) with respect to postpositioning. Whereas in the latter case post-

posing of the subject noun phrase occurs only after the gender copyingrule .has applied, this is not so with kuti clauses, since the kuti byitself has not the feature / J class/ which would make it possible forthe gender copying rule to apply. In their surface structure (53b) and(56c) resemble (5?b) and (58b) in respect of subject postpositioning.But it has already been pointed out that in some cases kuti clauses areunder the domination of NP as in (kO); whereas in others they are not,as in (5 0) - (52)* It is also observed that the kuti clauses which aredominated by NP take the concord zvi when the gender copying rule is

applied with them, as subject as in (55&) and (56b), The question whichimmediately suggests itself is, how do some of the kuti clauses come to

docontrol the concord zvi, while others^not?

If it Is maintained that NP may be expanded only into S, (an approach which however has been proved already to result in a very weak grammar) in, say, the string kuti udzokere kumusha1 in (55), then it is admitted that we have a situation in which an S dictates and controls

the concord zvi with the gender copying rule. What Is curious in this case is that the element which signifies the class feature zvi is absent or, shall we say, is not overtly expressed,. Elsewhere in the language noun phrases which take the concord zvi usually have themselves a zvi (or one of Its allomorphs) which is overtly expressed as in the sentences

in (59)*/ / /(59) a.* Zvipunu zviye zvataika.

(spoons - those - got lost _

Those spoons got lost.)/ s /b. Ndanzwa zvinhu zvinonakidza.

(I heard things which were interesting.)

3?A notable exceptions however* to this is a situation in which there is a combination of noun phrases which refer to non-human beings and which are in a subject or object position* Such a combination requires thiszvi for concordial agreement as illustrated in the sentences below*

/ / / / /(60)a, Mombe (10) nemadhongi (6) zvinofura pamwe chete,(cattle - and donkeys - graze - together =Cattle and donkeys graze together*)/ t / , , / * ' , S / /b, Bhuku rako (3) nepeni yako (9) ndazviisa patafura*

(book - your - and pen - your - I put them - on table =

As for your book and h e x your pen, I have put them on thetable,)

A stronger argument is however the following: the kuti clause has a sentential base as has already been pointed out. The kuti clause itself is derived by applying to the underlying S some transformational rules*

The result so obtained may be a kuti clause of the type in (40) which is dominated by NP, or in (30) which is dominated directly by VP, Whatwe have here then is a strange situation where actually one and thesame kuti clause is a noun phrase in one case and is not a noun phrase

in another as the following sentences illustrate,(61)a, Ndinoda*kuti auye,

(I want - that - he come = I want him to come,)/ / / // /b, Ndinozvida kuti auye*

(I want it - that -■ he come = I want it that he should come.)/ / / / / / / / /(62)a* Deni anofanira kuti auye,

(Deni ~ ought - that - he come = Deni ought to come,)/ / / s / / ///b, *Deni anozvifanira kuti auye.

(*Deni - ought it - that - he come,)The grammaticality of (61b) shows that the underlying structure of (6la)meets the structural description for object anaphora, hence the claim

that the string kuti auye is dominated by NP* Sentence (62b) indicates

38that this transformation blocks, and hence its ungrammatically and the

fact that kuti any/ is outside the domination of NP*It has already been said as well as demonstrated that the

sentences in (53) and (55)? tor instance, contain embedded sentences.That this is so can be illustrated further by listing the two sentences

which underlie each one of them. This can be seen in (63) and (6 4) res­pectively*

_ ~ » / / /( 6 3 ) Mwoyo anozviziva,(Mwoyo knows it,)/ / / / /b, Kondo akaba man,

(Kondo stole some money,)

(64)a* Zvinonetsa,(It is difficult,)

b, Unodzokera kumusha{(You will go.back home,)

(It should be noted however that the pairs of sentences in (63) and (64) are only approximate representations of the sentences which under­

lie (53) and (55) respectively*)It is observed that in these embedded sentences nowhere is

there any constraint on the noun phrases which they may contain. Further­more we see that the element zvi is present only in the higher or matrix sentence in both cases. But these zvi concords, as we may call them, areclearly reflexes of some noun in class 8, and among the noun classes inShona it is only class 8 which requires zvi as a concord. Therefore, inorder to account for the presence of these zvi concords or agreementsit is proposed here to postulate a noun of class 8 to be the head noun of such noun phrses as occur in the direct object position in (53) and in the subject position in (55)» It should be borne in mind nevertheless

that nowhere in these noun phrases with complement sentences do we ever get an actual noun of class 8 in the surface structure.

39The structure which underlies (33) may therefore b© represented

by the tree diagram in (63&) below. (Details which are not crucial to the present discussion are omitted in this branching tree. This will

also be the case in respect of the tree diagrams which will occur in the rest of this study.)

(63)eU

AUX PDF

I I IMwoyo Tns VP

no NP

zlva S.

NP AUX PDP+CL84-pro^

Kondo Tns

Iaka V

VP

NP

i iba mari

After complementizer placement has: applied we then get the structurein (63b) .

6. In this chapter complementizers are regarded as being introduced transformationally. The consequences of this approach are discussed in Chapter Four.

2*0(65)b

NP|

AUXI

IMwoyo1

Tns11no

PDP

zxva

PDPHP AUX

Kondo Tns VP

aka V NP

i lDa m a n

The structure which underlies ( 33) will have the representation in. (66a)* (66)a*

PDP

PDPNP AUX+N+CL8+pro

VPTnsxwe

no V

no V

netsa

daokera kumusha

As befores after complementizer placement has applied the structure in

(66b) results*

(66) b *. Sn

IP AUX PDP

!*■-MM: ^ IMF AUX FJUF no V+GL8 / + subi*/ / I I I(+P^O I I I Iiwe

dzokera kumusha/Iwe is a second person singular pronoun which in (55) is deleted. Haw-

©Ter, its deletion is optional;' sentence (67) in which, it is not deleted is perfectly well-formed,

... / / / ✓ ✓(67) Kuti iwe udzokere kumusha zvinonetsa.

In the account which, follows it is intended to deal with the complement sentences which are not covered by the expansion of IP into 1 + S, In this connection consider the following sets of sentences:

(Toni - wants - that — he go - to home =Toni wants to go home,)

/ / / i / / / /b, Kuti aende kumusha ndizvo zvaanoda Toni*(that - he go - to home - it is it - that he wants -- Toni

That he goes home is what Toni wants,)/ / / ✓ ✓ /c. Kuti aende kumusha ndizvo zvinodiwa naToni,

(that - he go - to home - it is it - that is wanted - by

Toni = That he should go home is what is wanted by Toni,)

(For' you to go back home is difficult*)

1-3 The: Case For Verb Phrase Complements

(68)a, Toni anoda kuti aende kumusha.

bz/ / / s , . / / / /.d. Toni anoda chii? (reply: kuti aende kumusha)

(Toni wants what? (reply: that he should go home))/ / / / . / / / /e» Toni anozvida, (referring to kuti aende kumusha)

(Toni wants it. (namely, that he should go home)), - . f / / / / / t / /(69)a* Toni anofanira kuti auye.

(Toni- ought - that - he come = Toni ought to come.)/ / / / / / / / /b* *Kuti auye Toni ndizvo zvaanofanira.// / / * / /c. *Kuti auye ndizvo zvinofanirwa naToni.

/ / / / / * ' * Sxd* *Toni anofanira chii? (reply: kuti auye)

e. *Toni anozvifanira. (referring to kuti auye)(70)a. Nd/ge yakaramba ichibhururuka.

(aeroplane - kept on - while it is flying =The aeroplane kept on flying.)

/ / / / / / /b, *Ichibhururuka ndege ndizvo zvayakaramba.c, *Ichibhururuka ndizvo zvakarambwa nendege.

/ / / / * . s / ^d, *Ndege yakaramba chii? (reply: ichibhururuka)

e, *Ndege yakazviramba. (referring to ichibhururuka)/ / / /The complement clause kuti aende kumusha in (68) exhibits the same pro­

perties as noun, phrases. For instance, it can be focussed in a pseudo­cleft sentence as in (68b); it can appear as subject in a passive sent-

/ence as in (68c); it can be replaced by the interrogative pro-form chii

as in (68d); and it can undergo object anaphora as in (68e). These pro-// / .perties are not observed with the complement sentences kuti auye in (69)

f tand ichibhururuka in (70).Although the complement clauses in (68) and (69) differ in

respect of their grammatical function, their derivation is similar.That is, their derivational histories share the same set of transform­ational rules as was pointed out on p.37 . The situation which we have

here is that, although (69) differs from either (68) or (70) in some

respects, it also resembles both of them in other respects. The trans-

43 / /formational rules which are required to generate ichibhururuka in (70)

differ considerably from those which are required in order to obtain

kuti auye in (69)®According to the way in which transformational theory is cur-

rently formulated there are a number of ways of characterizing these three sentences. One way is to regard the three complement clauses in(68), (69) and (70) all as instances of an S which is under the domina­tion of NP, but differring in the way the transformational rules whichapply to this S are specified. For example, we could say that when the

/ / /verbs -fanira (ought) and -ramba (keep on) are followed by an NP, that NP does not pass the tests for object, noun phrases. There would have to be yet another restriction with the verb -ramba which prevents the inser­tion of kuti before its S complement. All these restrictions make this particular approach needlessly complicated, and in any case there seems to be very little, if any, advantage to be gained by adopting it.

Another way is to treat the complement clauses in (68) and(69) together and (70) by itself. That the complement clauses in (68) and (70) show more differences than similarities seems to be obvious enough not to require discussion. The differences between the complement clauses of the type in. (68) and (69) were discussed at some length on p,3 0ff , and it was concluded;., there that our grammar becomes much

simpler and achieves a larger measure of descriptive adequacy if we treat them separately.

The third alternative postulates, that the complement clauses

in (69) &ud (70) are not instances of noun phrase complementation,, hut

rather that they are under the direct domination of VP* Our two verbs / / ✓-fanira and -ramba, instead of being marked with restrictions which

make the tests for object noun phrases fail, will be strictly subcate­

gorized in a way which allows themjto occur directly before S, and -da before NP, With this approach the ungrammaticality of some of the sent­

ences in (69) and; (?0) is automatically accounted for, because the tests for object noun phrases will only be relevant in the case of an S that is under the domination of NP,

ble hypotheses. For one thing, both approaches require a strict subca-

tegorization statement which is needlessly complex. For another, they

require restrictions on the transformations which are permitted. Needless to say, a statement of these restrictions would be very cumbersome in some cases. So it is the third alternative which is adopted for the

present.The phrase structure rules which are required for verb phrase

complementation are the same as in (A9) above except that

The structures which underlie the sentences in (69) and (70) may accord­ingly be represented by the tree diagrams in (71) and (72) respectively. The trees in (71e) and (72a) represent the deep structures of these sentences, while (71b) and (72b) are intermediate structures after com- plementizer placement has applied.

The first and second alternatives are less acceptable as via-

VP V(S)

(71)*

NP AUX PDP

Toni Tns VP

no V S’1

fanira NP AUX PDP

Toni Tns VP

(71)b.

NP

IToni

(72)a*

NP

Ndege

46(72)b

Ndege Tns

PDP

aka

0 / + partic_j7 NP AUX PDPramba

VPndege Tns

no

bhururuka

1*4 Sentential Complements and Selectional RestrictionsThat not every verb can occur with any noun phrase, either in

subject or in object position, in a sentence has been so often remarked

in linguistic literature that a detailed discussion in this connection is superfluous* The noun phrases have to obey the selectional restrict­ions that are imposed by the verbs* These selectional restrictions equally

apply to sentence complements which are dominated by NP* (For a detailed

discussion on selectional restrictions between verbs and noun phrases, see Chomsky, 1965a.) These selectional restrictions form in a way part of the meaning of a verb* As an example, consider the verb -nwa (drink). This verb takes as direct object a concrete noun phrase. If in the course of time it changed so that it could take as its direct object an abstract noun phrase, like rudo (love), it would no longer express the meaning which it has now. According to the meaning which -nwa'' expres­

ses now, (73a) is well-formed but not (73h).

(73)a <> Vana vanonwa mukaka.(Children drink milk.)

b 0 *Vama vanonwa rudo.

47(^Children drink love.)

Verbs which take direct object noun phrases will be markedas occurring in the environment NP, that is, they occur before anNP. If this environment is absent in the strict subcategorization state­ment of a given verb, that particular verb may not be followed by a direct object noun phrase. Besides, every verb will normally o c c u f in the environment NP » That is, every verb must have a subject nounphrase in a sentence as is indicated by (49a) above. We need not markthe environment NP since there are no exceptions to it* However,

in both the cases where the NP is in a direct object position and where

it is in a subject position it is necessary to specify what featuresthe NP may contain. As an illustration consider the sentenceSin (74).

t I *( 7 4 ) Hove dzinofema.(Fish breathe.)

b 0 Mudzimai akapona mwana*wake musikana.(woman - bore - a child - her - a girl =A woman gave birth to a baby girl.)

/ / /c. Nyaya yakona.

(The plan has failed.)/ / /d# Vanhu vazhinji vanotendera makuhwa.

(Many people believe gossip.)The selectional restrictions of the verbs in these four sentences in respect d»f the selection of the subject ' NP and the object NP can be represented in the following manner: __

(75)a, -fema + V+ </+ animate/^

+ Fn

-kona + V+ abstract/

+ Fn

-pona + V+ Z+ human7

d. -tendera + V+ Z+ humany

+ NP + NP+ +Z+ human7 + H'/J abstract^

♦*

+ Fn*

+Fn

The symbol </+ FnJ simply stands for the other features, syntactic or semantic, which may need to be specified* Notice that this is only a particial specification of the features of these verbs* Those features

which are not relevant to the present discussion have been omitted*In (7hc) and (7 fd) we have verbs which must take, respectively,

an abstract noun phrase as subject and an abstract noun phrase as direct object. With these abstract noun phrases there are only two possibilities

in regard to their structure* The head noun of such a phrase may be anoun which signifies an abstract idea like the following noun: makuhwa

✓ / / / /(gossip), mafambiro (a manner of walking or travelling), hasha (anger),rudo (love), ruvengcT (hatred), urombo'* (poverty) * The other possibility is that the NP may have an embedded complement clause* In the latter case the head noun in. Shona is not realized in the form of a noun as such in surface structure as has already been pointed out. Hence the following two sentences in (76) as well as those in (7*fc) and (74d)

are possible./ / / / / /(76)a* Kuti Toni atize zvakona.

(that « Toni - he run away - it has failed ssFor Toni to run away it has become impossible.)

/ / / / /b. Vanhu vazhinji vanotendera kuti kuna Mwari.(people - many - they believe - that - there is - God =

Many people believe that there is God.)

49But neither of the following sentences which contain respectively the verbs -kon.a and -tendera, and where the subject noun phrase and theobject noun phrase respectively are not abstract, is possible *

/ / / /(77)a * *Mvura yakona *(*The water/rain has failed*)

/ s /b* *Vanhu vazhinji vanotendera zvigaro.(*Many people believe chairs®)

Although, as was stated above, there are no identity constrai­

nts on the noun phrase in the complement sentences, there are nonetheless restrictions on the nature of the verb of the higher S® Thus, while the sentences in (78) are all grammatical, those in (79) with the same super­ficial structure as the corresponding sentence in (78) are not because of the nature of the verbs in the higher sentences* The complement clau­ses are underlined in the sentences below,

// / / s / / /(78)a . Kuti none chiva chichifamba zvinoshura*(that - you see » a puff-adder - it moving - it is ominous =

For you to see a puff-adder moving is an ill omen.)/ / / / / / / / / / / /b* VaNzungu vakaramba kuti rnwana wavo aroorwe naMberi*

is(Mr.Nzungu - refused - that - child - his - she. married -Aby Mberi = Mr,Kzungu refused to allow his daughter to marry Mberi.), A / , „ /. / /c, Amai vabvunza kana Farai akauya*(Mother asked whether Farai came.)

/ s / / sd, Midzi ati, "Ndinouya mangwana.1*(Midzi said, "I will come tomorrow,”)

. / / / / / ✓ ✓ /(79)a, *Kuti uone chiva chichifamba zvinorara.(*For you to see a puff-adder in motion sleeps.)

t / / / / ✓ / ✓ / / / /b, *VaNzungu vakabika kuti mwana wavo aroorwe naMberi.

(*Mr*Nzungu cooked that his daughter might marry Mberi.)/ / / / /c, *Amai vadya kana Farai akauya.

50

(*Mother ate whether Farai came.)/ / / / /d* *Midzi ataura, "Udinouya mangwana."

(*Midzi spoke, "I will come tomorrow."The verb must be one that can take an abstract noun phrase as either subject (for verbs which behave like -shura (serve as ill omen),

"finha (get bored), -dzimaidza (confuse), -era (be sacred, taboo),-fumura (put to shame), -ipa (be bad), -naka (be good) and many others)or object (for verbs which behave like -bvunza (ask), -bvurn.a (agree),

/ / / /-yeuka (remember), ~da (want, need), -ti (say), -edza (try), -fembera(guess) and many others)..

What has been established in this chapter is that some complement clauses function as noun phrases, hence such complement clauses are said to be dominated by NP* As noun phrases they may ap­

pear in subject or object position. The complement clauses which do not function as noun phrases are directly dominated by VP.. The complement clause (clauses) which may occur in any one sentence is governed by the verb in the higher sentence..

2*0 COMPLEMENTIZING MORPHEMES 51I n this chapter I shall give a tentative scheme of the comple­

mentizers which are found in Shona* I say tentative because the issue is far from being settled. What seems to be clear however is that there is need to set up a grammatical category of complementizers* This chapter will be devoted to arguing the case for complementizers.

2*1 Kuti /+ indicative7 and Kuti / + sub.junctive7Let us examine the underlined; complement clauses in (ZfOa) and

(78b) in Chapter One which are repeated here as (1) and (2) for ease of

reference.i / / / / / / / /(1) Mwoyo anoziva kuti Kondo akaba mari*

, N / / / ✓ / / / / / / /(2) VaNzungu vakaramba kuti mwana wavo aroorwe naMberi*We notice that the formative which is performing the function of subord­

inating the constituent clause to the matrix sentence is kuti. Neverthe­less, kuti in (1) is followed by a clause in the indicative mood,, while kuti in (2) is followed by a clause in the subjunctive mood* This diffe­rence in the mood of the clause which follows kuti is dependent to some

extent on the verb in the containing S* As will be seen later the mood of the constituent clause forms part of the subcategorizational state­ment of the verb in the higher S* The two sentences given above provide us with two distinct complementizers, namely, (a) kuti indicative? and (b) kuti / + subjunctive?* Further examples of sentences in which the constituent clauses are introduced by these two complementizers; are given

below*Examples with kuti indicative?

, v s / / / / / / / / /(3) Kuti Nixon akanyepa zvakasvota vanhu vekuAmerica*(that - Nixon - lied - it angered - people - of at America - That Nixon told a lie angered the Americans*)

f / i f f f /( k) Kuti murume nemukadzi vanombotukana zvinozivikanwa*

(that - a man - and a woman. - sometimes; quarrel - it is known =That a man and his wife sometimes quarrel is known.)

52, X t / f t t(5) Ndino f unga kuti zvinhu zvichanaka munyika muno.

(I - think - that - things - will he all right - in country -this = I think that the situation will improve in this country,)

* i f f / / /(6) Podzorimwa abvuraa kuti akatadza*(Podzorimwa admitted that he made a mistake*)

Examples with kuti subjunctive/, % / / / / / / / / / /(7) Kuti ambuya vanwe tii isina mukaka zvava kuramba*

(for - grandmother - she drink - tea - it not having - milk -

it is now - refusing = For grandmother to take tea without milk is now difficult*)

// ✓ / / / S / / / / / /(8) Kuti muroora aende kunoona amai vake zvakarurama*(for - daughter-in-law - she go - to see - mother - her - itis all right - For the daughter-in-law to go and see her mother

is in order*). . / / / ✓ . / / /(9) Vakomama avo varangana kuti vatrze*

(boys - those - planned - that - they run away - Those boys

have planned to run away*)

2*2 Kui: / + infinitive/On p*38 it was postulated that the head noun of noun phrases

with complement clauses is and this works very well with the

complement clauses which we have seen so far and which are under thepN

domination of NP* Nonetheless: not all complement clauses have +CL8 as1+prqj

their head noun. With some complement clauses there is need, it seems,+N

to postulate something like +CL15+Pro

as the head noun. An example of

this is the complement clause in (10) below*i ' / / ✓ / / ■(10) Vana vadiki vanoda kutamba bhora*

(Little children like to play football.)

To illustrate that there is need to postulate something other than +CL8jfcpro

for the head noun of the complement clause in (10) consider the applica­

tion to it of the following transformations in (11):

53(11)a* pseudo-cleft transformation;

i f f s S s f fKutamba bhora ndiko kwavanoda vana vadiki.(to play - football - it is it. - which they like - children - little » Playing football is what little children like*)

b* passive transformation;;f t ! t f t f t sKutamba bhora kunodiwa nevana vadiki*

(to play - football - it is liked - by children - little = Playing football is liked by little children*)

c* object anaphora transformation;/ / / / / / / /Kutamba bhora vana vadiki vanokuda*

(to play - football - children - little - they like it =sAs for playing football e little children like it*)

/ / /What these transformations tell us is that the string kutamba bhora isfunctioning as a direct object noun phrase* But clearly this noun phrase

* / /contains a verb phrase, namely, -tamba bhora. This suggests that the noun phrase has a complement clause embedded in it. That the string-tamba bhora is part of an embedded sentence is illustrated by the fol­

lowing transformations in (12)./ / / / f / / /(12)a. Bhora rinoda kutambwa nevana vadiki.

(football - wants - to be played - by children - little =It is small children who should play football.)

/ f f t f ( /b. Vana vadiki vanoda kuritamba.(Small children like to play it..)

i f t ( / / / /c. Vana vadiki bhora ndiro ravanoda kutamba.(children - little - football - it is it - which they like-

to play = As: for small children football is the gamewhich they like to play.)

In (12a) the complement clause has undergone passiviaation, in (12b) it

has undergone object anaphora, and in (12c) it has undergone pseudo­cleaving. These transformations apply within: a simplex sentence.

Now notice that the concordial agreements which are controlled

by the string kutamba bhora in (11a) through (11c) are quite different

54from; those which we saw controlled by kuti Kondo akaba mari in (40b) in Chapter One* Whereas the complement clause in (40b) has the concordial

agreement zvi, that in (10)- has ku*. The former is an affix of class 8, while the latter is an affix of class 15* It is worth noting also that the complement clauses which are dominated by NP in Shona control norm­ally either class. 8 or class 15 agreements* There are nevertheless in­stances of complement clauses which, though in subject position, do not

control either class 8 or 15 concordial agreements, at least in surface structure* Such are the complement clauses in the sentences below.

(13) Kuti Jojo akabirwa mari^hachisi chokwadi*(That George had his money stolen is not. true*)

(14) Kuti murume anyatsogara nevakadzi vaviri haisi nyore*(For a man to live peacefully with two wives is not easy.)

(15) Kuti mtwana^ asvike* makore maviri asati^ afamba hacliimbori^/ / /chinhu chinotyisa*

(For a child to go for two years before it walks is not something to worry about*)

Notice that in these sentences the subject concord is controlled, not by the subject of the sentence as one would expect, but by the. comple­ment noun phrase which is underlined. This is a feature which is charac­teristic of the verb "to be" in Shona.

It is also possible however for the subject complement clause

in. (1 3 ) - (1 5 ) to control the subject prefix in the usual manner as isillustrated in (16) - (18) respectively.

/ / / ✓ / / / f(16) Kuti Jojo akabirwa mari hazvisi chokwadi*/ / / / / / / / / f /(i?) Kuti murume anyatsogara nevakadzi vaviri hazvisi nyore.

V / / i / / / / ✓ ✓(18) Kuti mwana asvike makore maviri asati afamba hazvimbori

chinhu chino tyisa-.The connotativ.e difference between the sentences in (1.3) - (15) oil the one hand and. those in (1.6) - (18) onthe other seems to be that attent­

ion is focussed on the complement noun phrase in the former sentences,

and on the subject NP in the latter sentences.. Notice that (18) may

55

alternatively occur as (19) below but with the same semantic, reading.i t t i t / / / / / / / /(19) Kuti mwana asvike makore maviri asati afamba hazvimbori

/ , / szvinhu zvinotyisa.K £The. change from the singular chinhu chinotyisa to the plural formi

/ /zvinhu. zvinotyisa is only in form, not in meaning* In this case it isnot clear whether the subject concord is controlled by the subject con-

/ /plement clause or by the complement, noun phrase zvinhu zvinotyisa..The. generation of (13) - (15) presents a difficulty which

however is more apparent than real.. Given the gender copying rule.(see 5*-7) the sentences in (16) - (18) are easily derived. That is,

there is no problem im describing them. But in (13) ~ (15) there is need to describe how the complement,, and not the subject, comes to con­trol. the subject prefix* Notice (a) that the control, by the verb comple­ment of the subject prefix is optional in these sentences and (b) that this phenomenon occurs in a structure of the type

NP1 - V - NP2

where V is the verb "to be". This apparent discrepancy can be obviated by setting up another rule of gender copying which is optional and which applies, not from left to right, but from right to left*. This rule will apply Just in case the verb is "to be"..

Control of a subject prefix by a complement clause is not re­stricted to sentences with subjects which consist of embedded clauses only as is illustrated below.

(20) Gadheni rangu haisi kirechi*.(My garden is not a creche.)

f t / f t / ✓(21) Mombe hachimbori chinhu cheku tamba nacho.(Cattle are not things to play with.)

Compare (20) - (21) with (22) - (23) In which the subject controls the subject concord* The meaning remains the same.

(22) Gadheni rangu harxsi kirechi., £ f t / / / /(23) Mombe hadzimbori zvinhu zvekutamba nazvo*

Further examples of sentences in which the complement clauses in subject

56

position do not seem to control either class 8 or class 13 concords are:. . / / / ✓ / / /(2%) Kuti Jeso akatifira ichokwadi.

(That Jesus died for us is true.), N / / / / ✓ / / / /(25) Kuti ungazvidziirire nemazwi akadaro kutamba zvako.

(that - yon can defend, yourself - with words - like that -

it is playing - yours = To think that you can defend yourselfwith those words is child’s play..)

, / / / / / /(26) Kuti haasiriye akaba mari inherna.((The denial) that he is not the one who stole the money is false.)

In a very few cases complement clauses may also control class 16 con­

cords. These will be discussed In 6.1*5.Apart from showing that the head noun of a complement clause

under the domination of HP need not always be a pronoun of class 8, thecomplement clause in (10) appears to offer us another complementizer,namely, ku.. This ku is followed always by a verb in the infinitive

f i tmood.. But since the complement clause kutamba bhora controls ku agree­ments, one might ask whether these ku agreements in (11) are not merely reflexes of the ku in kutamba. In other words, is ku not a noun classconcord like zvi in avipunn. in (59a) Chapter One? In that case the

/ / Jstring kutaimba bhora would be regarded as something like a derived nominal. However, the sentences in (12) above do not support this view. Derived nominals in Shona normally do not undergo the transformations which we see in (12). There will be more discussion concerning derived nominals in the next chapter. ^

Furthermore, we might want to examine the following sentence .;/ s f t / /(27) Peni anyatsa kuvhura gonhi.

(Peni - did carefully - to open - door =- Peni opened the door carefully.)

We observe that kuvhura gonhi is not a case of noun phrase complement­

ation since the tests for object noun phrases fail here as is shown

below in (2 8).

y?

(2 8)a. pseudo-cleaving test;/ / / / /*Peni kuvhura gonhi ndiko kwaanyatsa.

b, passive test;/ t A /* Kuvhura gonhi kwanyatswa naPeni*

c» interrogative pro-form test;/ / / / , / .*Peni anyatsa chii? (reply: kuvhura gonhi)

d, object anaphora test;^Pen^akunjratsa. (referring to kuvhura gonhi)

This is a case of verb phrase complementation (cp* (69a) in Chapter IE). No adverbials between -nyatsa and kuvhura are possible* Although the string kuvhura gonhi is a sentence constituent which is directly domina­ted by VP in the containing sentence, its own VP consists of a verb accompanied by a direct object noun phrase, namely, gonhi* Notice also how similar in their surface structure (10) and (2?) are. The configura­

tions which underlie these two sentences are presented in (29) and (3 0) respectively.

(29)

vana vadiki Tns

no

da S1

NP AUX PDP

vana vadiki Tns VP

tamba bhora

vhura gonhiApart from lexical items and grammatical formatives these two underlyingstructures differ only in that in (29) the complement clause is dominatedhy NP, while in (30) it is dominated by VP. What is nevertheless moreimportant to realise is that the set of transformational rules which map

f t finto kutamba bhora in (2$) seems to be the same as the set of rulesthat are required to transform in (30) into kuvhura gonhi» If thatis the case, then ku is merely a complementizer in both instances andconsequently a pronoun of class 13 has to be postulated in the underlyingstructure in (29), but not in (3 0 ), in order to account for the kuconcord agreements we see in (11).

Now, the realization of the class 13 concord shares the same phono-p N ”*

logical form with the complementizer ku. Just as + CL8+pro

does not receive

phonological realization in surface structure when it appears as a headnoun of a complement clause as was. pointed out on p. 3& > so also does

never get realized phonologically in surface structure+N+ CL13 +pro

59when it appears as a head noun of a complement clause. This view receives

further support from a consideration of the sentences in (31)*•' t t / t i / /(3 1 )n* Mako anoda kufamba nevasikana.

(Mark - loves - to walk - with girls = Mark loves to he in

the company of girls,)i / / / / / / /, /b, Mako kufamba nevasikana anokuda,

(Mark - to walk - with girls - he loves it = As for movingin the company of girls, Mark loves it.)/ / / / ■ / / / / /c, Mako kufamba nevasikana anozvida.

(Mark - to walk - with girls - he loves it = As for movingin the company of girls, Mark loves it,)

The underlined ku in (31b) is an anaphoric pronoun of the complement k

clause kufamba nevasikana. The element zvi in (31c) is also an anaphorict f / s fpronoun of the same: complement clause kufamba nevasikana. But the senten­

ces in (31b) and (31c) carry the same meaning. Therefore in this casethe complement clause: may take either ku or zvi agreements. If it is

/ / / / /maintained that the ku in. kufamba nevasikana is the feature which expres­ses the noun class, we. are hard put to it to explain the presence of azvi agreement in (31c). But if, on the other hand, we say that in its

/ s / / tunderlying structure kufamba nevasikana has a pronoun of class 15 as its

head noun in (31b)> and a pronoun of class 8 as its head noun in (31c), then this- problem will not arise. The implication here is that there is a choice in this case with respect to the head noun of the complement clause. But clearly we have not this choice of head noun with regard to

the complement clause in (40b) in Chapter One. There the concord which is controlled by the complement clause is zvi and none other. That being the case- there is need to specify when this choice can be exercised and when it cannot. This choice may tentatively be stated in the following

manner: when the complement clause is preceded by the complementizer ku

the head noun may be a pronoun of either class 8 or class 1 5 , with the latter being more often preferred to the former; but if the complement-

60izer is any other than ku, then the head noun is restricted to class 8 only.

2.3 KutiFor the present I would like to examine more closely the com-

plementizer kuti. It would appear that in some cases the -ti in it is averb, and moreover a verb which can take a direct object noun phrase* Tothis end consider the following sentences:

i f f / / t / /(32)a. Deni anyatsa kuti Toni abve pano.(Deni - did clearly - to tell - Toni - he go away from - here

=Deni told Tony clearly to go away from here.)/ / / / f i t /b. Toni anyatsa kunzi naDeni abve pano.

(Tony - did clearly - to be said - by Deni - he go away from -here » Tony was told clearly by Deni to go away from here.)/ / / / / / s /(33)a. Garwe anoda kuti vanhu vamuite marabo.

(Garwe - wants - that - people - they make him - a chief =Garwe wants the people to make him their chief.)

/ / , / / / / / /b. Kunzi nevanhu aite mambo Garwe anozvrda*(to be said - by people - he become - a chief - Gar vie likes it

= Garwe likes it that the people make him their chief.)

The complement clauses in (32) and (33) are instances of complement clauses under the direct domination of VP and NP respectively. In. (32b) and (33b) -ti has been changed to its passive form -nzi. Notice that al­though (32b) and (33b) contain the passive form of -ti, only (32b) is synonymous with its corresponding active sentence (32a). In both senten­ces in (32), because -nyatsa does not express an action or a state as verbs normally do, but is rather expressive of adverbiality, it is not a true verb semantically. It is a modal-like verb. This presumably exp­

lains why the pair of sentences in (32) are paraphrases. However, struc­

turally these sentences seem not to be related. Although the passive

readily applies in in (32b*) the further movement of Toni to be the

subject of SQ is strictly speaking not part of the passive transformation.

61

/In (33^) the verb of the higher S is a true verb -da. In this sentence it is Garwe himself who wants to bring it about that the people make him their chief. But in (33b) Garwe would welcome a move by the people to

make him their chiefs but he may not have the means to bring this about. To this extent (33a) and (33b) are not synonymous. The point to note how­ever is that in both (32) and (33) -ti is a fully fledged verb and not a complementizer. Hereunder are the tree diagrams which show the structures which underlie the sentences in (32) and (33)* These trees help to show the differences between the members of each pair which have

been mentioned above.(32a1)

Nl

1

nyatsa NP VP

Deni V

ti 32

+N+CL8+pro

NP VP

Toni V NP

bva pano

62

NP

■Toni

VP

SV 1

nyatsa NP

Toni

,VP

PP \ 2

NP

ha Peni

bva pano

In of (32b ) first object raising applied to make Toni the direct object of the verb -ti then the passive rule applied next and lastly the rule of extraposition*

(33a1)

NP1

GarV/e V11da 1

VP

NP

ti

¥N+CL8+pro

t- -vanhu

ita Garwe mumbo

Gar we

aa

+N+ CL8 +pro Garwe

nzi

na vanhu V NPI Iita mambo

(33b11)

VPNPNP

Garwe

daNP+N+CL8+pro

PPGarwe

VPnzi

na vanhu V

11 a mambo

The transformational rules applying to in (33k ) are the same as1 1 those which applied to in (32b ) above,, Then in (33k ) SQ undergoes

the rule of Object Focussing (see which moves the direct objectof ~dsf to the left of the subject noun phrase leaving a pro-form in itsplace as in (33k11)* It is from this intermediate structure in (33k11)that the sentence in (33k) is finally derived.

64-

In the case of (32'b) we observe that the complement clause kunzi naDeni abve pano continues to be under the direct domination of VP. This is evidenced by the fact that this embedded clause does not control concordial agreement anywhere in the sentence as would be expected of a complement clause which is functioning as a subject. Moreover if we apply the second test for subject noun phrases to (32b) we come up withToni as the subject noun phrase of this sentence.

f t / i / i t / / f(34.) Ndiani anyatsa kunzi naDeni abve pano? (reply: Toni)(it was who - who did clearly - to be said - by Deni - he goaway from - here = Who was clearly told by Deni to go awayfrom here?)

We cannot use the second test in such a way as to elicit kunzi naDeni*1 i / as the subject noun phrase of (32b) because of its positionin the sentence, namely, that it is a verb phrase complement.

n ^ / / /On the other hand, the string kunzi Mevanhu aite mambo in (33b)/ / /controls the concordial agreement zvi in anozvida. This suggests that

[+Nas: the head noun of this string. Besides,there is something like +CL8

+prothis complement clause passes the second test for object noun phrases.

(33) Garwe anoda chii? (reply: kunzi nevanhu aite mambo)(What does Garwe like?)

What all this, indicates is that the status or function of the complement clauses in (32b) and (33b) remains the same whether -ti is passivized or not. That is to say, these complement clauses continue to be dominated by VP and NP respectively. If -ti is a verb in these sentences it follows that kuti cannot be regarded as a mere complementizer always. It is more complex than that. The following sentences demonstrate further the com­plexity of kuti.

f t / / / / / / / f(36)a. Musikana uyu anoda kuti, "Ndiuye mangwana here?1'(This girl wants to say, "Should I come tomorrow?"

/ / / / / / / / / /Musikana uyu anoda kuti ndiuye mangwana here?

(Does this girl want to bring it about that I come tomorrow?)

65

t t / / / // ✓ / /c„ Musikana uyu anoda kuti ndiuye mangwana here?

(Does this girl want me^ to come tomorrow?)

The meaning of -ti in these three sentences is not the same althoughthese sentences are identical in their linear arrangement of the words. In (36a) -ti means "say, utter", in (36b) it has the reading "bring it about that, make, cause", and in (36c) it seems to be a mere complemen­

tizer and the emphasis in this sentence appears to be on "me". So it would appear that there are at least three different uses of kuti and I should like, to discuss each one of them in some detail as they are crucial to the study of sentential complements in Shora*

2.3.1 Kuti (to say)The instances where -ti means "say"1 seem to be quits clear

and straightforward. I give a few examples of them in (37)../ / / i i / / /(37)a* Akamyara kuti aiva nemo ta muhapwa.(he was ashamed - to say - he was - with an abscess - in armpit = He was shy of saying that he had an abscess in

his armpit.)/ / / ' ✓ / / /b. Mashiri akabvotomoka kuti Keni akapusa.

(Mashiri - spoke unwisely - to say - Keni - he is foolish = Mashiri said without thinking about it that Keni was

foolish..)i l l / / / / / / /c*. Mukomana anobvuma kuti shamwari yake yakatiza.

(the boy - admits - to say - friend - his - ran away - The boy admits that his friend ran away.)

/ / / . / / id.. Sami anoramba kuti nyika yaka tender era.(Sami refuses to admit that the world is round.)

e*. Tsuro akatdnga kuti Bvemi aie numgo.(The hare began to say that the baboon was lazy.)

All. these sentences in (37) are amalyzable in a similar manner. Taking

(37d) as an example, we can represent its underlying structure by the

diagram in (3 8).,

66(38)a<

S ani

¥PNP

V NP

ti/+ 0 coiujd/

+pronyika

tenderera

(38)b,

Sani

ramba

ku/ T in finj7

ti N S

0 comp NP

nyika V

tenderera

67

1Applying 0 complementizer insertion and gender copying in the second

cycle, and ku / +" infinitive/ insertion and Equi-NP deletion in the third

cycle in (38a), the result is the tree in (38b), (Most of the transform­ational rules which are being mentioned now will be discussed In greater detail in Chapter Five, Notice also that as in Chapter One the trans­

formational hypothesis of complementizer insertion is adopted in this

chapter*) From this derived phrase marker is eventually obtained the sentence in (37d)*

But if after applying 0 complementizer insertion and gender copying, object raising is then applied in the second cycle of the con­

figuration in (38a) the result is the tree in (38c).(33)c*

NP VP

Sani V NP

ramba ku ZJ in fin..U

VPNP

Sani V NP S

nyika 0 comp VP

V

tenderera

To this, tree can be applied the object focussing rule in the second cycle, and then the same rules, in the third cycle as were applied above* The structure obtained is something like (38d). After applying later rules

1* This 0 complementizer is discussed in 2.9 in this chapter.

ramba N

VP+N+c l 8*pro

NP ku| /+ infiru7

nyika V Sp

ti 0 comp1

VP

V

tenderera

and the rules of the phonological component we eventually come up with

Notice that object focussing is a common phenomenon in Shona as the

sentences in (40) illustrate»

(I have eaten sadza already,)/ /ii» Sadza ndadya kare.

(Sadza - X have eaten - already = As for sadza I have

already had it.)/ / / / / / / / /b.i. Sekuru anoda kuenda kumusha asi mari haana.

(uncle - wants - to go - to home - but - money - he has

none - Uncle wants to go home but he has not the money.)

(39)«(39) Sani anoramba nyika kuti yakatenderera.

(Sani refuses to accept that the earth is round.)

Ndadya sadza kare

69/ / / / / / / /ii, Kuemda kuniusha seku.ru aitoda asi mari haana.

(to go - home - uncle - wants - but - money - he has none =As for going home uncle likes it but he has not the money,)

That nyika is still part of is shown by the fact that the stringnyika kuti can undergo anaphoric pronominalination as in (4 1),

/ / / / / / /(hi) Sani anoramba kuiti yakatenderera,(Sani refuses to admit that it is round,)

Since anaphoric pronouns of the type exemplified by i in (%I.) occur'almost always only before verbs in the language, it can safely beconcluded that -ti in (37) is a verb.

Similarly the underlying structure of (42a) below may be transformed into the structure which eventually yields (42b), The

sequence of rule application is the same as for (41) except that, in place of anaphoric, pronominalization, reflexivization will apply forits structural description will have been met,

/ n * ( i f f '(42)a, Mujubheki akati ava murungu,(Mujubheki said that he had become a white man,)

t / / / / /b, Mujiubheki akazviti ava murungu*(Mujubheki said of himself that he. had become a white man,)

Observe also that this -ti can be passivized as in (43b),

(See. also (32) above,.)/ / / / / / /(43)a, Toni aramba kuti Deni akapusa.

(Toni refused to say that Deni is foolish,)f t / f t / /b,. Deni aramba kunzi naToni akapusa,

(Deni refused to be said by Toni that he is foolish,)/

In. (^3a) Toni, the subject noun phrase, is followed by an active verb-ramba which in turn is followed by a complement clause containing an

/active verb also -ti. The subject noun phrase Deni in (43b) is also fol­lowed by the active verb -ramba which is itself followed by a complement clause with a passive verb. In this case the mere fact that there are

two different subject noun phrases but the same active verb occupying the

same position in these two sentences indicates that these sentences can­not have the same semantic reading. Admittedly this may only he contin-

Kgently true. Nonetheless, although these two sentences are not paraphra­ses of each other, it is clear that -ti has assumed a passive form in (A-3h)s and passivization is one of the phenomena which are attributed

to verbs. It seems evident from the foregoing discussion that -ti in

(36a) is a fully fledged verb.

2.3*2 Kuti (to bring it about)I would like now to examine the second kuti which appears in

(36b). The -ti in kuti can also be demonstrated to be a verb. To this

end consider the following sentence:(AA-) Taka/ anoda kuti vana vafunde.

Sentence (A f) is in itself ambiguous. On one reading the kuti has the meaning which it has in (36b), and on another it has the meaning which it has in (36c). Here I am concerned with the former reading. The under­

lying structure of ( k k ) appears in (A5a) below.

(Taka - wants - to make - children - they learn

Taka wants the children to learn.)

(45)a.

Taka V NP

+N NP+CL8 I

VP

£ P r2J Taka V NP

ti

~1 ■ NP+CL8

VP

V

fun da

71

Applying kuti Zi subjunctive/ insertion and gender copying in the second cycle, and ku /+“ infinitive/ insertion and Equi-NP deletion in the third cycle of the structure in the result is the configuration in. (Jf5b).

(^5)b<

Taka

ku ^ in fin./

+ nr o

'ti S2VPNP

vana V

funda

From this rough structure is finally obtained the sentence in (kk)*But if, in addition to applying kuti /F subjunctive/ insertion

and gender copying in the second cycle, object raising is also applied, the result is the structure in (£f5c)»

1 72

Taka

+ ku I * in fit i j p/ jT n+CL8+pro

Taka Sp

VPkutiE su'd1 7 I

VIfunda

Ag was done in the case of the other kuti above the abject focussing rule can apply now but still in the second cycle* In the third cycle the same rules as before are applied., yielding the structure in

Taka

da

VPNP

vana

ti VPkutiE sub 3^7 j

V

funda

73Front this structure is eventually derived the sentence in (Zf6) •„ i f / i f f t(4 6 ) Taka anoda vana kuti vafunde.

(Taka - wants - children - to make it - they learn « Taka wants to make the children learn,)

Notice that when the complementizer formative kuti occurs after the verb -ti it is obligatorily deleted in surface structure* Since S^ in (45c)

meets the structural description for the rule of anaphoric pronominali-zation this rule can be applied resulting ultimately in (47)*

i f i i i t f(if7) Taka anoda kuyati vafunde, (referring to vana)(Taka wants to make them learn*)

We observe also that the -ti in this kuti can assume a passiveform as shown in (4 8)*

t i f f i f f(48) Vana vanoda kunzi vafunde naTaka,(the children - want - to be made - they learn. - by Taka =

The children are wanted by Taka to learn,)Once more all these transformations seem to confirm that the -ti in this second kuti is. a verb* The structural descriptions of these transforma­tions which have been discussed above normally require the presence of

a verb. Hereunder are further examples of this use of kuti* The sentences in (i) - (iv) in each of (49a)s (i+9t>) and (49c) correspond to thesentences in (44) s (46) 9 (47) and (4 8) respectively,

/ / / / f / /(49)a. i* Mambo abvuma kuti vanhu vake vaite kanzuru,/ / i f f t i tii. Mambo abvuma vanhu vake kuti vaite kanzuru,

(The chief has agreed to allow his people to form alocal council,)

f i / ( i /iii, Mambo abvuma kuvati vaite kanzuru,(The chief has agreed to allow them to form a localcouncil,)

i f f t aiv* Vanhu vabvuma kunzi vaite kanzuru namambo,

(the people - agreed - to be made - they form - council - by the chief = The people have agreed to form a local council at the order of their chief,)

?kf t i t / / / / /

Id. i. AmlDuya vakaramba kuti mwana aende kuSmoia*/ / / / * t / / / / .ii. Ambuya vakaramba mwana kuti aende kuSmoia,

(Grandmother refused to let the child go to Sinoia.)iii. Ambuya vakaramba kumutx aende kuSxnoxa.

(Grandmother refused to let him go to Sinoia.)iv. Mwana akaramba kunzi naambuya a e n d e kuSinoia.

(the child - refused - to be made - by grandmother - he go - to Sinoia = The child refused to be ordered by hisgrandmother to go to Sinoia.)

t / / / i /c* i. Mufudzi anogona kuti mombe dziuye masikati./ / / / / /ii. Mufudzi anogona mombe kuti dziuye masikati.

(the herdsman - can make - the cows - they come - in

the afternoon = The herdsman can bring the cows in the

afternoon.)/ / / ✓ / / /xii. Mufudzi anogona kudziti dziuye masikati.

(The herdsman can bring them in the afternoon.) iv. Mombe dzinogona kunzi dziuye masikati nemufudzl.

(The cows can be made to come in the afternoon by the herdsman.)

From the discussion so far the following facts have emerged: (a) that there are at least two different verbs' both with the phonological shape -ti, one has the meaning "say, declare" and the other has the meaning "make, bring it about that, cause;"; (b) that these two verbs differ also

in their rule feature specification. The former requires that its com­

plement clause be in the indicative mood as in (37) or a quotation as in (36a), while the latter requires its complement clause to be in the subjunctive mood.

2.3*3 Kuti (a complementizer)I now turn to the kuti in (36c). I shall illustrate the syntac­

tic function of this kuti by taking the sentence in (A4) which resembles

(36c) in its structure. The only syntactic difference between the two

75is that whereas (36c) is a question (A4) is a declarative sentence and therefore slightly less complex in its derivational history* It has al­ready "been pointed out that (AA-) is ambiguous* Here I am concerned with

the second reading of this sentence (see p,^° )* I repeat this sentence below as (50) for convenience,

(50) Taka anoda kuti vana vafunde*(Taka wants the children to learn.)

Notice that although (50) and (51) differ from each other in their linearsequence, they are nevertheless paraphrases of each other.

t / / / / / /(51) Taka anoda vana kuti vafunde,(Taka wants the children to learn.)

In (A?) above we saw an anaphoric pronoun occurring before -ti. Anaphoricpronominalization in this case was conditioned by the verb -ti. Thesentence in (5 0) can also appear with an anaphoric pronoun but this time

;in the position immediately before the verb -da as in (52),(52) Taka anovada kuti vafunde,

(Taka wants them to learn.)In this case anaphoric pronominalization is governed by the verb -da!

Let us assume, for arguments sake, that the structure which underlies

(50) is the same as the one which is outlined in (A5a) which is repeated here for convenience as (53)*

(53)

Txaka

da/ku + infi .1

'+N +CL8 +pro

NP VP

Taka V*tti/ F u t i + s u b J7 Vvana

VPfunda

76

Adopting this analysis for (50) raises a number of problem's* Firstly, there does not seem, to be any sound transformational machinery which can raise vana, the subject noun phrase of S2 > through S^ to become the: direct object noun phrase of the verb -da in SQ * Notice that in (51) vana/ is a direct object noun phrase of -da in SQ otherwise anaphoric pronominalization would not have applied in (5 2)- In the configuration in (45c) vana was raised from a subject position in S0 to be a direct object noun phrase of -ti in S- , Sentence (52) would require this same vansf to be raised again from this direct object position in to be a direct object of -da/ in the, next higher sentence, namely, Sq* Secondly, the two sentences, viz* (17) and (52)j in which the raised NP vana^ occurs as an anaphoric pronoun in each case have two different semantic

interpretations. If both these sentences are derived from an underlying structure like that in (53) Ly applying a series of different transform­ational rules, it then follows that the difference in meaning between these two sentences is attributed to the transformational rules applied.

The verb -ti will be the same and it will impose kuti / + subjunctive/ insertion on the complement clause in. both cases, (As was stated before complementizers are introduced transformationally in this chapter.) It is only the raising of vans'* to two different positions in this structure to which the difference in meaning can be ascribed. This approach clearly admits that transformations are not meaning-preserving and this is con­trary to one current theory in transformational grammar, and the oneadopted in this study, namely, that transformations do not change the

2meaning of the structure to which they are applied. If there could not be found an alternative way of accounting for such data within this current theory, then the theory itself would be open to question for

there are quite a number of verbs in. Shona which give us the same

patterning.. Further examples are given in (54) ** (56)-

2- For a discussion on whether transformations alter the meaning of the structures to which they are applied see Barbara Hall Partee "On therequirement that transformations preserve meaning'1 in Studies inLinguistic Semantics. C..J*Fillmore and D.T.Langendoen (eds-)

77/ / / e / /(5Zf)a. Ndagona kuti Shambare agare pasi.

(X managed to make Shambare sit down,) b* Ndagona kumuti agare^ pasif (referring to Shambare)

(I managed to make him sit down,)/ / / ( / /e, Ndamugona kuti agare pasi. (referring to Shambare)

(I prevailed upon him to sit down.)/ / / / / / / / / / / I(55)a® Mudzimai uyu akatadza kuti murume wake arege hwahwa.

(This woman failed to persuade her husband to stop drinking beer.)

t f / / t / t /b, Mudzimai uyu akatadza kumuti arege hwahwa. (referring to/ / / fmurume wake

(This woman failed to make him stop drinking beer.)/ / / / / / / / / .c* Mudzimai uyu akamutadza kuti arege hwahwa. (referring to

/ / / ✓murume wake(This woman did not prevail upon him to make him stop

drinking beer-), _ / / / , / ✓ / / / / /. / / ./ / /(56)a, Murungu akatendera kuti Goto aende kunoona amax vake vaxrwara.

(The boss allov/ed Goto to go and see his mother who was ill.)/ / / / / / / / / / / / . / / /b* Murungu akatendera kumuti aende kunoona amax vake vairwara.

(referring to Goto)(The boss agreed to make him go and see his mother who was

ill.)/ / / / s / s s / / f / / f fc. Murungu akamutendera kuti aende kunoona amax vake vaxrwara,

/(referring to Goto)(The boss allowed him to go and see his mother who was ill.)

1 believe that the structure which underlies (5 2) is different

from that which underlies (47)• la other words the diagram in (53)is not the underlying structure of (52).. The claim made here is that kuti in (5 2) i& a complementizer, not a verb, and that this complement­izer may be followed by a clause in the subjunctive mood or in the indi­

cative mood. It is proposed that the underlying structure of (52) is as

78outlined in (5?a) below*

(57)a.

Taka

§uti £ F subj.,7 / ^object raising/

N S

VP+c l 8+pro

vana V

funda

No transformational rules of interest to us in the present discussion apply in the first cycle. In the second cycle kuti £+ subjunctiv£7

insertion applies resulting in the tree structure in (5?b)*

(57)b.

Taka

daobj, r a i s ^

N S1

vana

VP

V

funda

79If we choose not to apply any of the transformations determined hy the optional rule features which are governed by the verb -da, the resulting

sentence will be (5 0)*But if, after applyng kuti /J* subjunctive? insertion and

gender copying as above, we choose to apply object raising, still in the second cycle, to the tree in (57b) we get the structure in (57c).

(57)c.So

NP VP

Taka

funda

If no other (optional) rule features of -da are applied we eventually

obtain the sentence in (58)*(58) Taka^ anoda vana kuti vafunde,

(Taka wants the children to learn*)In this sentence the focus is on van/. Notice that sentence (5 8 ) is iden­tical, in terras of linear sequence, with sentence (if6)« Nevertheless their transformational histories are different. In (58) van/ is a direct object noun phrase of -da whereas in (if6) It is a direct object noun phrase of -ti which has been front-shifted in order to focus attention on it. Now if to the structure in (57c) we apply anaphoric pronominali­zation we get the sentence in (52)* Here are further illustrative exam-r:-~ pies but with complementizer kuti indicative? this time. The sentences (i) - (iii) in each case correspond to the sentences In (50), (58) and

kuti

80(52) respectively*

/ f f t / / / t(59)a.i, Mako akamzwa kuti Tembeni akatiziswa,/ / / / / / /ii, Mako akanzwa Tembeni kuti akatiziswa.

(Mark heard that Tembeni eloped with someone*)/ / / / i t / /iii. Mako akamunzwa kuti akatiziswa.

(Mark heard that she eloped with someone.)t t i ( f f t /b.iW Wdinosiva kuti Maria naKembo vanodana./ / / / / / /ii. Ndinoziva Maria naKembo kuti vanodana.

(I know that Mary and Kera.bo love each other.)/ t / . / / /iii. Mdinovaziva kuti vanodana.

(I know them that they love each other.)t i t t / /c.i. Zuze ataura kuti nyaya yacho yapera*/ t / , , /ii* Zuze ataura nyaya yacho kuti yapera.

(Zuze said that the affair was over.)

iii. Zuze aitaura kuti yapera.(Zuze said that it was over.)

t i / / / / ./d.i. Tande acherechedza kuti Keni haana kubvisa m a n *t i / / / t / ./ii, Tande acherechedza Keni kuti haana kubvisa man*(Tande noticed that Keni did not pay any money.)t t / t t / / t ./iii* Tande amucherechedza kuti haana kubvisa m a n *

(Tande noticed it of him that he did not pay any money.)/ / / / / ✓ // z' /e.i* Baba aidaira kuti roora richabviswa*/ ✓ / / / / / ^ /ii, Baba aidaira roora kuti richabviswa,

(Father believed that the bride price would be paid.)/ / / / / / / / /iii, Baba airidaira kuti richabviswa.

(Father believed that it would be paid.)t / /f.i. Tinotenda kuti mombe dzaita basa,/ / /ii, Tinotenda mombe kuti dzaita basa*

(We are thankful for the work that the oxen have done.)t / / /iii. Tinodzitenda kuti dzaita basa*

(We are thankful for the work they have done.)

81One further observation with this use of kuti as a complementizer is that after object raising has applied the passive rule can be applied with the; resulting sentence being synonymous with the active sentence*.This presumably is due to the fact thajs passivization is operating in the topmost S. The following are the passives of the sentences in

(58) and (59)*(60)a* Vana vanodiwa naTaka" kuti vafunde. (cp. (58))

(The children are required by Taka to learn.)/ / t / f ✓ / ✓ / , V Vb. Tembeni akamzwikwa naMako kuti akatiziswa. (cp. (59a))

(Tembeni - was heard - by Mark - that - she eloped with

someone = Mark heard that Tembeni eloped with someone.)/ / f / / / f t S t ! ,c. Maria naKembo vanozivikanwa neni kuti vanodana. (cp, (59b))

(Mary and Kembo are known by me to love each other.)A t f f f /d*. Nyaya yacho yataurwa naZuze kuti yapera, (cp. (59c))

(The affair was said by Zuze to be over.)f t t f t ! te. Keni acherechedzwa naTande. kuti haana kubvisa mari. (cp.(59d))

(Keni was noticed by Tande not to have paid any money.)i f / / / / f f * sf. Roora raidairwa nababa kuti richabviswa, (cp. (59e))

(The bride price was believed by father to be about to be paid.)

/ / f / /g. Mombe dzinotendwa nesu kuti dzaita basa. (cp. (59f))

(the oxen - are, thanked - by us - that - they did work =We appreciate the service done by the oxen for us.)

But when kuti was turned into the passive in its other two uses which were discussed above, the active sentences and their seemingly corres­ponding passive sentences did not carry the same meaning. This was thecase v/ith the pair of sentences in (k3)> and the sentences in (k4) and( 8).

Notice also that this kuti does not seem to function as a verb anywhere. If it is insisted that -ti here is a verb, then this -ti, unlike the other two that have been seen, does not permit object raising

82

and anaphoric pronominalization as the sentences in (61) indicate* These

sentences correspond to those in (59)*(61)a. *Mako akanzwa kumuti akatiziswa* (referring to Tembeni)

/ / / / / / f / zb. *Ndinoziva kuvati vanodana* (referring to Maria naKembo)^i t s * . / , . / / .c. *Zuze. ataura kuiti yapera* (referring to nyaya yacho)t t * / / / f .d. *Tande acherechedza kumuti haana kubvisa tnari. (referring to

/Keni)/ / ✓ / / / Z' s , 'V ^e» *Baba aidaira kuriti richabviswa. (referring to roora)/ / /xf* ^Tinotenda kudziti dzaita basa. (referring to mombe)

However, a verb which is followed by a noun phrase as part of its strict subcategorization need not allow object raising (if the NP is abstract

and contains an embedded sentence) or anaphoric pronominalization. But in the case of this kuti if we maintain that it contains a verb -ti, we are hard put to it to describe in a convincing way how the subject noun, phrase of its complement clause can move the whole way up to be a direct object noun phrase of the main verb of the sentence which domina­tes this kuti as was pointed out on p* 76 .

/But as we have already noticed with the verb -da. anaphoric/pronominalization can apply in the VP that immediately dominates -da

itself (see (52)) or in the VP that dominates -ti (see (if?))* Neverthe­less it has also been pointed out that the two sentences which we get have different semantic interpretations. So -da happens to be a verb

which can allow both ku /_± infinitive!/ insertion and kuti £ ± subjunctive/ insertion. Most of the verbs in (61) on the other hand, while permitting kuti / T ~indicative/ insertion, do not allow ku / V infinitive/ insertion.

It is also observed that the passive sentences in (60) can all occur with kunzi substituted for kuti without effecting any further change either in the structure or in the meaning of these sentences*

3* Sentence (61b) would be acceptable if -ti meant "say".

83

This is shown in- (6 2)*t t t t / / /(6.2)a* Vana vanodiwa naticha kunzi vafunde*/ / / / j / / /b* Tembeni akanzwikwa maMako kunzi akatiziswa*

/ / / / / / / / / /g *, Maria naKembo vanozivikanwa meni kunzi vanodana*

/ ✓ / / / /d* Nyaya yacho yataurwa naZuze kunzi yapera*f t / / / / /e* Keni acherechedzwa naTande kunzi haana kubvisa marl.

/ / / / / / / ✓ / / /f* Roora raidairwa nababa kunzi richabviswa*/ / / //g* Mombe dzinotendwa nesu kunzi dzaita basa.,

It has already been stated above that kunzi is the 'passive form of kuti

when kuti is a verb as in (43) (48)* Recall also that normally whena verb is turned into the passive form; in a sentence, that sentence undergoes a structural change, the most noticeable change being that the object noun phrase becomes the grammatical subject of the passive sente­nce* In the sentences in (60) and (62) we have then a strange situation in which kuti, an active form of a verb, interchanges freely with its passive form kunzi without effecting any structural change in the sente­nce* We also have the strange situation in which kunzi in one case is a

true passive which causes the normal structural change as in (43)(4 8), and in another case is a passive which has no syntactic function.

2*3.4 Kunzi

There are strong indications that kunzi in (6 2) is not a true passive verb* For one thing it appears to be one of a very few verbs in Shona (Zezuru dialect) which behave in this curious fashion* Notice also

that kunzi in (6 2) occurs only after a verb which itself is in the pas­

sive form in the higher sentence*, But in (43^) and (48) the verb in the higher S is in the active form*, This seems to be the crucial difference. I believe that this kunzi is a complementizer, just as kuti with which it appears as a free variant in such sentences is. Its passive form can be described as some form of assimilation which is conditioned by the preceding passive verb* This assimilation however does not seem

8if

to go through more than one S* It is possible that diachronically the complementizer kuti/kunzi might have been the infinitive form of some active verb "ti (presumably a different -ti from the other two that were discussed earlier on) and its passive form -mzi, but that it lost its verbal character in the course of time. Synchronically, however, this kuti/kunzi is a mere complementizer. This phenomenon is by no means restricted to -ti only. There are other verbs in Shona which have either lost or are in the process of losing their verbal character and assuming a different categorial status. In this regard consider the following two verbs:, -nyanga (although, even if) and -bva (go eewy, begin at, come

from:)*As a true verb -nyanga may be found in sentences like the

following:/ / A / / i f / /(63)a. Sara, nechirwere chake ichi, anyangoenda kuchipatara/ / / / /hazvichambomubatsira.

(As for Sara’s illness, even if she goes to the hospital it will not help.)

/ / / /b* Kunyangoendako hapana chaunowana.(Even if you go there there !©• nothing that you will get.)

In these sentences -nyanga Is an intranstive verb which cannot passivize under any circumstances. Notice that in (63h) this verb is In the infi­nitive mood. But there is another infinitive form in which the final vowel is /e/ as in kunyange (although, even if). This second infinitive form can occur with the passive extension as in kunyangwe with exactly the same meaning and without causing any structural change In the S in which it appears. That the meaning of -nyanga and kunyange/kunyangwe is the same is obvious^ enough.. But kunysuage/kunyangwe is now functioning

as a conjunction of some kind or possibly as a complementizer as shown

in (6^)./ t / f t /(6A) Kunyange akauya nhasi hazvibatsire.

(Even if he comes today it won’t help.)

85

Furthermore, although -nyanga as a true verb (synchronically at least) requires that its complement clause be a contracted infinitive, kunyange/

kunyangwe on the other hand is followed, not by a contracted infinitive^, but by a clause which is in the participial mood or consecutive mood as

in (65) and (66) respectively,/ / / / / / / / /(65) Nzou inogona kusimudza danda kunyange rakakura sei.

(An elephant can lift up a log no matter how big it may be,)(66) Sara^ ne^chirwere chak/ ichi, kunyange akaenda kuchipatara

/ / / /hazvimubatsire•(Sara, as for her illness, even if she goes to the hospital it will not help her,)

In fact kunyange can be followed also by an NP or by a relative clause

as in these two sentences:/ / ( / / /(67) Handidi chinhu nhasi kunyange tii zvayo.

(I do not want anything today, not even tea,)1 f f t f t f(68) Nzou inogona kusimudza danda kunyange rakakura chaizvo,

(An elephant can lift up a log even one which is very big,) Moreover the forms kunyange/kunyangwe can occur simply as nyange/nyangwe with the same meaning, Nyange/nyangwe cannot in any way be regarded as imperatives as one would expect. The infinitive and the imperative forms of this verb now perform the same function, namely, that of a conduction

of some sort.The verb -bva would occur in a sentence like the following:/ / / / / / /(69) Tande akabva pano svondo rakapera.

(Tande left this place last week.)

(70) Tande aida^ kubva pano svondo rakapera.(Tande wanted to leave this place last week.)

In (70) this verb appears in the infinitive mood. But the infinitive form kubva can also mean simply "since, from that time" which in a way

A* Contracted infinitives are discussed in 6*2.1,2.

86

is still related to the meaning of -bva given on p.8if c With the meaning "since" kubva is a conjunctive of some kind. Its syntactic behaviour is different from that in (7 0).

, , / / / / i f f / /(7D Suzi ari kungorwara kubva musi uye,(Suzi has been ill since that day.)

That kubva in this case derives from a verb is shown by the fact that

it can occur with the applied extension as in (72). The meaning of the sentence remains unaltered.

. t t t ( / / / / /(72) Suzi ari kungorwara kubvira musi uye*(Suzi has been ill since that day.)

Extensions such as the applied, the passive and so forth are features which are normally associated only with verbs in Shona.

2.A Kana /+ consecutive7For the moment let us consider the embedded sentences in the

following two sentences.✓ , t , t , f f(73) Kana amax vaToko vakamutuka hazvibatsire*

(Even if Toko's mother reprimand her it won't help*)f / / /(7if) Kaaa akasvika pano zvingashure.

(If ever he gets here it will be an ill omen.)The embedded clauses which are underlined are NP complements, in subject position* Both of them pass the tests for noun phrases In subject posi­

tion. In the first place, both strings kana amai vaTokcf vakamutuka and kana akasvika pano control the concordial agreement zvi in their respec­tive sentences. (This concord zvi is underlined.) Secondly, if the interrogative pro-form test is applied the questions which result arethose in (75) and (76) respectively.

, . / / / / /(75) Ghii chisingabatsire?(What won't help?)

/ / /(76) Ghii chingashure?

(What would be ominous?)

8?/ ✓ / ✓ / .The replies to these questions are kana amai vaToko vakamutuka (even if

Toko's mother reprimand her) and kana akasvika pano ( if ever he gets here) respectively.

Both of these NP complements: are introduced by the word kana« while the clause so introduced is in the consecutive mood. (Notice that the word "consecutive" as used here serves only as a label referring to the tone pattern on the verb, and not necessarily to consecutive action semantically.) In the noun phrase complements of the type exemplified

in (73) and (74) we are provided with another complementizer, namely, kana /+“ consecutive/,

2.3 Kana / T participial/There is yet another complemetizer which is introduced by the

word kana. In this case the clause following kana is in the participial

mood as- in (77) - (80).(77) Kana Pedzi auya nemarx yedoro zvanaka,

(If Pedzi has brought some money with which to buy beer it is all right,)/ / / / / i(78) Bomba haachayeuka kana makambosangana,

(Bomba no longer remembers whether you have met before.)t / s /(79) Ndakanganwa kana ndakazvxreva,

(I can't remember if I said it,), _ . / / / / / / / / / / /(80) Feso akatanga kurangarira kana aizoona nyika yake zvakare,

(Feso began to wonder whether he would see his fatherland again.)

In (77) the complement clause is in subject position while in (78) - (80) the complemet clauses are in. object position. It can be demonstrated

easily that these complements are in effect noun phrases. In (77) thef ✓ / / / /string kana Pedzi auya nemari yedoro controls the concordial agreement

/ /zv in zvanaka. Furthermore the interrogative pro-form test yields the string

88

(81) Chii chahaka?(What is all right?)

f f / t f fto which the answer is kana Pedzi auya nemari .yedoro (if Pedzi hasbrought some money with which to buy beer)* To illustrate that the com­

plements in (78) - (8 0) are object noun phrases I shall, consider (78) only. It passes all the tests for object noun phrases*

(8 2) a*, pseudo-cleaving;-Kana lirak arab 0 s an g an a ndizvo/zvaasisayeuke Bo^mba.(Whether you met before is what Bomba no longer remembers*)

b* passive;f t y f f ^ fKana makambosangana hazvichayeukwa naBomba*

(whether - you met before - is not remembered - by Bomba =Bomba does not remember whether you have met before*)

c. interrogative pro-form;/ f / / / / / NBomba haachayeuka chii? reply: kana makambosangana)

(Bomba does not remember what? (reply: whether you have

met before))d* object anaphora;

/ / / / / / /Bomba haachazviyeuka* (referring to kana makambosangana)(Bomba does not remember it*)

2*5*1 Among the clauses which are introduced by kana /+ participial/ are embedded or indirect questions such as those in (7 8) - (80)* A brief comment on non-embedded questions is in order here since it will help us to understand the nature of embedded questions in Shona* The main interrogative formatives in Shona are: herfe. (yes/no); -pi (which one?); -i (what sort of?); and/(who?); rinhi (when?); sei (how, in what way?); -ngaiod/ (how many?) and -su (expressing dissatisfaction.)*. These areunderlined in the examples below.

/ / /(8 5 ) Vana vadya sadza here?

(Have the children had their meal already?)

89

(8 4) Mombe dzaenda kupi?(Where have the cattle gone?)

(85) Waona oh±± kumunda?(What did you see at the field?)

.... / / / /(86) Sara akarwa naani?(Sara fought with who?)

/ / /(87) Chitima chinosvika rinhi?(When will the train arrive?)

/ / / ' /(88) Jongwe akaita izvi sei?(Why did Jongwe do this?)

These interrogative formatives are mutually exclusive in any one simple

sentence* The interrogative formative in (83) however differs from the others* The question which it signals requires either a "yes'1 or a "no" for an answer, whereas the questions in (8 4) - (88) require much more information than Just a "yes" or a "no”* The answers to the questions

in (8 4) - (87) are noun phrases of one kind or another, The interrogative/ / /formatives -pi, -i, ani, rinhi and ngani can in effect substitute for

/noun phrases. The question in which sei occurs requires an answer which states the way in which a given action (or state) is accomplished (or entered into). Question formation in Shona is a topic which is as yet to be investigated more fully. The brief comment above on this topic

will nevertheless be adequate for the purpose at hand,how

2.5*2 Now let us consider^the question in (83) above differs from

(89).(89) Sara‘S abvunza kana vana/ vadya sadza,

(Sara asked whether the children have had their meal.)

Notice that (8 9) resembles in structure the sentences in (78) - (80). Although compelling evidence is still scanty, it would seem that the string &ana vana/ vadya sadza is in effect a question which requires a

"yes11 or a "no" for an answer in the same way that (83) does. In this

90

respect it is especially interesting that (89) may he paraphrase^ by(90) in which the "yes/no" question actually appears.

(90) Sara /bvunza mubvunzo unoti, "Van/ vadya sadza here?"(Sara asked the question, "Have the children had their meal?"

Another point which has already been noted is that the string kana vana^✓vadya sadza is a noun phrase complement*

/It is observed also that the interrogative formative here

occurs in non-embedded questions only, while kana in this particular case occurs in embedded questions* That is, the two are in complementary distribution. They seen* to constitute one morpheme in this case* A string

like the one in (91) is totally ungrammatical,(91) *Sara abvunza kana vana* vadya sadza herel

Another interesting observation is that kana introduces an indirectquestion of the here type only and none of the others as is shown bythe ill-forraedness of (92) - (96)®

/ / / / / /(92) *Sara abvunza kana mombe dzaenda kupi*(93) -'Sara abvunza kana waona chi/kumunda®(94) *Sekai abvunza kana Sara* aliarwa naanif(95) ‘*Sara abvunza kana chitima chinosvika rinhi*

/ / / / / / / /(96) *Sara abvunza kana Jongwe akaita izvi sei.For these othei” questions kuti / + indicative/ is the introducing comple­mentizer as shown by the grammaticality of (97) - (101),

/ / / / / /(97) Sara abvunza kuti mombe dzaenda kupi*; (Sara asked where the cattle went.)

/ / / / / /(98) Sara abvunza kuti waona chii kumunda.(Sara asked what you saw at the field.)

/ / / / / / /(99) Sekai abvunza kuti Sara akarwa naanx.(Sekai asked who Sara fought with.)

/ / / / y(100) Sara abvunza kuti chitima chinosvika rinhi.(Sara asked when the train would arrive.)

91f f f f / f J /(101) Sara abvunza kuti Jongwe akaita izvi sei*

(Sara asked why Jongwe did this.)J /Notice that kuti £+ indicative can also introduce a here type of ques­

tion but only if the formative here'' is retained in a position after the verb of the clause which it questions as in (102)**

(102) Sara'' abvunza kuti vana^ vadya sadza her el(Sara asked whether the children have had their meal.)

Sentence (102) is synonymous with (89)* Other examples of sentences like(102) are given below*

/ / / / / /(103) Ndinoda kuziva kuti makama mombe kare here*(X want to know whether you have already milked the cows.)

(104) Xichaona kuti Mber/ anogona kutamba bhora herel(We shall see whether Mberi knows how to play football,)

, . / / / / ' / / s * / /(105) Shamwari yangu ichandiudza kuti munhu uye akadzoka here.(My friend will tell me whether that person came back.)

All the sentences in (103) - (105) may occur alternatively with kana / T participial/ as complementizer for their embedded questions without effecting any semantic change.

(1 0 6) Ndinoda/kuziv/ kana makama mombe kare*(107) Tichaona kana Mber/achigona kutamba bhora.(108) Shamwar/yangu ichandiudza kana munhu uye Zcadzoka.

2.6 Deletion of kana

In this connection there are four occurrences of kana to be considered which differ according to their meaning* These are kana (timewhen), kana (if, expressing a condition), kana (although, even if), andkana (whether). The rest of this subsection will be devoted to describing w^en kana may be deleted and when it may not.

2.6*1 kana (time when)

This kana is invariably followed by a clause in the participial

92

mood. If the tense in the higher B is either a (recent past) or aka (remote past), kana is obligatorily deleted as shown by the ill-formed- ness of the (a) sentences and the well-formedness of the (b) sentences

below,/ * t y(109)9-* *Kana vasvika vakagara pasi,

/■ / / / / /b, Vasvika vakagara pasi,(On arrival they sat down,)

/ / s / t / f / /(110)a. *Kana vavhimi vaehipedza kudya vakapinda munzira.f t / f t / / / /b, Vavhimi vaehipedza kudya vakapinda munzira.

(Having finished their meal the hunters set out on their

journey,)/ / / / / / / /(111)a, *ICana mombe yaneta nekurima yakatanga kurara,

b, Mombe yaneta nekurima yakatanga kurara,

(When the ox was tired because of ploughing it began, to

lie down,)(112)a, *Kana vachaenda7 havafn& kunzw/ nzara asi pakudzoka.

/ / / t i f t / /b, Vachaenda havana kunzwa nzara asi pakudzoka,

(When they were still on their v/ay there they didn't feel hungry, but on their r wajr back (they did))

But if the tense indicator in the higher S is any other than a or a k a ^kana is optionally deleted without effecting any change in meaning,

/ / / . /(113)a, Kana vasvika vaigara pasi,b, Vasvika vaigara pasi,

(Having arrived they used to sit down,)/ / / / / / t f / / / f t(llif)a, Kana mvura youya hapana chakanaka nekupisa kwakuri kuita.

/ / / / f t i f / i i f fb D Mvura youya hapana chakanaka nekupisa kwakuri kuita,(When the rain comes it will be terrible because the weatheris very hot,)

f / / /(1 1 3)9 , Kana aguta hwahvra anotaurisa.t / f tb, Aguta hwahwa anotaurisa.

93

(When he has had a lot of "beer he talks too much*)/ / / '(116)a® Kana madya mosimuka moita basa*

b® Madya mosimuka moita basa(When you have had your meal you stand up and do some worj(.)

2.6*2 kana (if)

When kana introduces a conditional clause, that clause willbe in either the consecutive mood or the participial mood. Kana followed

by the consecutive mood may be substituted freely with kuti. This kuti

seems to be different from the other forms of kuti which were discussed above.

s / / / / /f /(117) Kana/kuti akagona kugadzira midzimu yese iyi anganzi murume*

This kana/kuti introducing a clause in the consecutive mood may be dele­ted optionally without disturbing the meaning of the clause® Hereunder

are the sentences (1 1 7) - (119) with kana/kuti deleted.

Examples of conditional clauses introduced by kana / T partic^J are provided below. -

(If Mavis didn't get here she didn't do the right thing.), * / / / / / / / / / / / /(124) Kana akafa achizvxsungxrira pamusana pemazwx enyu muchamuripa

/chete.(If he died by hanging himself because of what you said you

will eventually have to propitiate his spirit®)

(If he manages to propitiate all his ancestral spirits, he may be said to be a man.)

(118) Kana/kuti ukasvika pano ndinokurova® (If you come here I will beat you.)

/ / / ' s f t(119) Kana/kuti Musarapasi akaita izvozvo imhosva yake.(If Musarapasi does that it will be his fault.)

f t i t(120) Akagona kugadzira midzimu yake(121) Ukasvika J®ano ndinokurova.

/ t . . / < fyese xyx anganzx murume

(123) Kana Mavis asina kuzosvika kuno akatadza

94/ / / / s / y(125) Kana takanganisa tiudzeiwo*

(If we have done wrong please tell us*)In (123) " (125) kana may not be deleted as is shown by the ±11-formed-

ness of (126) - (1 28)*(126) *Mavis asinsf kuzosvika akatadza.f y y y / / / / y f t / /(12?) *Akafa achizvisungirira pamusana pemazwi enyu muchamuripa

/chete.

(128) ^Takanganisa^ tiudzeiwo*Notice that (125) could also mean "when we have made a mistake please

tell us11* With this meaning, that is, expressing "time when’1, kana is optionally deletable as was observed above, hence (1 2 8) would be a well- formed sentence in that case.

2*6*3 kana (although, even if)

With this meaning kana interchanges freely with kunyange and is normally accompanied by a clause in the consecutive mood. In this

case kana may not be deleted,/ / / / / /(129) Kana/kunyange ukashupika sei handimbokubatsira,

(No matter what difficulties you will get into I will not

help you*)/ y y / / /(130) Kana/kunyange kukanaya tinototi tiende chete nhasi kumusha.

(Even if it rains we will have to go home today all the same.) It would appear that kana/kunyange may be followed also by a clause in

the subjunctive mood as (131) and (132) show*/ / / y / / y y(131) Kana/kunyange ushupike sei handimbokubatsira,

/ / / / / / / / /(132) Kana/kunyange kunaye tinototi tiende chete nhasi kumusha.The sentences in (131) and (132) have the same meaning as (129) and (130) respectively. In (131) and (132) however kana/kunyange seems to beoptionally deletable.

f t / t / / / ;(133) Ushupike sei handimbokubatsira.

95t / / s / / ( /(134) Kunaye tinototi tiende chete nhasi kumusha*

But in the following sentence kana/kunyange may not be deleted without

either drastically altering its meaning or rendering the sentence ungram­matical*

/ / / / ' / f(135) Kana/kunyange akauya nhasi hazvichatiyamura*(Even if he comes today it : is not going to help us*)

If kana/kunyange is deleted from (135) the result is (136)*/ / / - / / / f(136) * Akauya nhasi kxxKXKhnfexkaksxrxjr hazvichatiyamura*

which as a paraphrase of (135) is ungrammatical* Sentence (136) can only

be well-formed if it expresses, not a concession as in (135)» "but acondition as in (137)* (cp the sentences in subsection 2,6*2)

/ ✓ ✓ * / / /(137) Akauya nhasi hazvichatiyamura,(If he comes today it will no longer help us.)

It is not clear to me at the moment why (133) a*id (134) should be well- formed, while (1 3 6) is not, yet they have undergone the same deletion rule,

2,6*4 kana (whether)

When Kana carries the meaning ’’whether" it introduces an

embedded question (see 2,5*2 above.)* In such sentences kana is not

optionally deletable as the ungrammaticality of (138) - (1 40) whichcorrespond to (106) - (108) respectively indicate.

/ / / / /(138) *Ndinoda kuziva makama mombe kare,

(139) *Tichaona Mberi achigona kutamba bhora./ / / / / / / / / /(140) *Shamwari yangu ichandiudza munhu uye akadzoka.

2.7 Zva / + ~participial7Zva appears, to be one of the complementizer formatives in

Shona* To this end consider the sentence in (141) below,✓ / / ✓ ✓ * ’ / / /(141) Temba zvaauya nhasi kudai zvinoratidza kuti basa akapedza.

(Because Temba has come today it means that he finished the

work.)

96/ / / / /■The string Temba zvaauya nhasi kudai seems to be a noun phrase comple­

ment, This string is the subject of the predicate zvinoratidza kuti basa / / / rakapedza. The zvi in zvinoratidza is dictated and controlled by this

noun phrase complement. Up to now it would appear that I have only assu­med without proving it that the string Temba zvaauya nhasi kudai is anoun phrase complement. In the first place, if the interrogative pro-form

/chii is substituted for this string in the sentence in (1A1) the resultis the grammatical question

✓ i / / /f(1*1-2) Chii chinoratidza kuti basa akapedza?(What indicates that he finished the work?)

/ // / /to which the answer is Temba zvaauya nhasi kudai. As was seen in 1,2*1 *.2the answer to a question with interrogative chii is a noun phrase. Sec­

ondly, the pseudo-cleaving of (1^1) yields (1A3) which is perfectly grammatical,

(lZf3) Temba zvaauya nhasi kudai ndizvo zvinoratidza kuti basa akapedza,(Temba's arrival today is what indicates that he finished

the work.)

2V0 in fldizvo is clearly a reflex of the prefix of class 8 which in this case is referring to the string Temba zvaauya nhasi kudai. Further examples of sentences: of this type are given below. The noun phrasecomplements in question are underlined.

/ / / / / ' / / / f(lh*f) Tayi zvarapera zvichaita kuti ndipinde muchikwereti.(Since this tyre is worn out I shall be forced to enter into a debt,)

(1A5) Zvawatora mari^yake zvichakukonzera kutaura,(Taking his money as you have done will lead you into a

quarrel with him,) ruie uyu zvaatever*

(By tracing his wife this man has shov/n that he loves her.)(li|-6) Murume uyu zvaatevera zvinoreva kuti anoda mukadzi wake".

97i t / / t / t /(1 4 7) Zvavakatenga ge.io zvakavabatsira kurima.

(Buying the plough helped them to till their land,)(148) Zvawava kuehovha makasa zvichakurunzira kuba.

(Since you are now gambling for money you will be led into stealing*)

Each of these noun phrase complements is introduced by zva. This zva is then invariably attached to the verb of the clause which it serves to introduce. The verb which follows it is always in the participial mood* Thus zva /+“participial/ is another complementizer. The meaning carried by zva in this case is "since, as, because — M.

2.8 Deleted Complementizer FormativesIn Shona there are some sentences in which the complement

clausejis not introduced by an overt complementizer such as kuti or kana as- was the case in the sentences which we have met so far in this chapter, In the sentences referred to presently the complement clause, may be in the participial, indicative, subjunctive or consecutive mood* All these seem to be cases: of verb phrase complements, not noun phrase complements.

Examples of sentences in which the complement clause is in the participial mood:(149) Mungadai maenda kumusha.

(You should have gone home.)(150) Vanhu vakaramba vachitaura vanzi nyararai^ namambo.

(The people kept on talking when they had been told to keepquiet by the chief.)

/ / / / / / / / /(151) Baba akadzamara asvrka zvese zvekutamba zvapera.

(As soon as father arrives all playing has got to stop.)(152) Tsitsi^ ndinogara ndichxsangana naye kutsime.

(As for Tsitsi I always meet her at the well.)(153) Jeke haachafa^ akazviit a7 zvakare gyekubal

(Jack will never steal again.)

98

An example of a sentence in which the complement clause is in the indi­

cative mood:

(15k) Ita pashoma nepashoma unodzamara wav/ana zvauri kuda.(Do it little by little; in the end you will get what you

want.)Examples : with complement clauses in the subjunctive mood:

(155) Amai vakadakara vadzoke kutsime Dambudzo asati amuka.(Mother returned from the well only to find Dambudzo stillin bed.)

(156) Zuze ndiye asaka tsuro ipunyuke.(Zuze was the one who let the hare escape.)

An example with a complement clause in the consecutive mood:, N / f t / / / / f /(157) Midzi akasimbirira akadakara akawana fundo yaaida.

(Midzi persevered until he got the education he wanted.)There are at least two possible ways of coping with the com­

plement clauses in (lk9) ™ (157)* One approach is to say that if these embedded clauses are introduced by any complementizers at all, then the complementizers in these sentences are: 0 //"participial/ for (lk9) -

(1 5 3), 0 //indicative/ for (15k) > 0 Z F subjunctive/ for (155) - (156), and 0 //" consecutive/ for (1 57).

However it is also possible that diachronically there were

overt complementizer formatives like kana, kuti and ku in these comple­mentizers but that these were eroded in the course of time so that today only the mood with which they were associated remains. Since all these

four moods have already appeared above with overt complementizer forma­tives one could postulate the following complementizers for these sentences:

kana [ V participial/ for (lk9) - (153);kuti £± indicative/ for (15k))

kuti / f ~ subjunctive/ for (1 5 5) - (156);

99kana { T consecutive/ f or (157)•

As has already been stated these complementizers will form part of the strict subeategorization of the verbs after which they occur. Just as some verbs, for instance, are marked for optional object deletion, the verbs in question, here will be marked for obligatory complementizer formative deletion.

If the latter suggestion proves to be viable, it is- to be preferred to the former which seems to proliferate the number of comple­

mentizers to be found in Shona. Nevertheless more investigation is requi­red before a more positive statement can be made in this connection. For the purpose at hand the latter approach is adopted.

2,9 Verb -ti (say) and ComplementizersDirect speech following on the verb -ti (say) - the only verb

in Shona privileged to introduce a quotation - requires comment. Such a quotation is to be regarded as a direct object of some kind. Here isan example of such a sentence.

/ i f / / /(158) Sara ati, "Toko uri rombe."(Sara said, "Toko, you are stupid,"

Although the complement in this sentence is not introduced by an overt

complementizer formative, this sentence differs from those in (1^9) *-(157) in that the complement is a noun phrase complement. If this sent­ence is transformed into the passive, and extraposition applied subse­quently,it will become apparent that the head noun of the quotation is a pronoun of class 8,

(159) Zvanzi naSara*, "Toko^ uri^rombe,"(It was said by Sara, "Toko, you are stupid."

Zv- is clearly a subject concord of class 8*Notice that -ti may also be complemented bjr an ideophone.

/ t / s // / ✓ S ^(160) Vana vakati zii kunyarara zvakapinda mudzidzisi.(The children became very quiet when the teacher came in.)

100

111 this case however -ti does not seem to mean "say, speak", but rather

something like "become".Since the quotation which may occur after -ti may be in almost

any of the moods, we cannot postulate, it seems, a complementizer forma­tive like kuti or kana before it, because doing so would lead us to include among the complementizers such impossible complementizers as:

kuti /J" imperative/, kana ZF indicative/ or kana Z F subjunctive/* In brief it would not be known in some cases which complementizer formative to postulate. In order to obviate this difficulty it is proposed here to set up tentatively a 0 complementizer which will form part of the

strict subcategorization statement of -ti (say) only*. This 0 complement­izer will serve to introduce a quotation.

2.10 List of Complementizers:Among the complementizers which are found in Shona (Zez.uru

dialect) are the following:kuti ZF indicative/ kaua/kuti ZF participial/kuti Z F subjunctive/ kana ZF consecutive/ku Z+~ infinitive/ zva ZF" participial/0 complementizer.

Although words like kubvira, dakara, dzamara and a few others introduce subordinate clauses, they nevertheless, do not seem to subcate-

goriz.e vdrbs, For this reason one hesitates to regard them as complemen­tizer formatives. They may be mere conjunctives, I have not included them, among my complementizers because their status is not clear to me

at the moment.It is proposed to regard each complementizer as being made up

of two parts, namely, the introducing element and the mood. A complement clause must be in one or another of the moods in Shona. The complement­izer in any complement clause is governed by the verb in the higher

101

sentence. Another observation worth noting is that complementizers are mutually exclusive. If one is inserted in any one cycle none of the others may be inserted subsequently in that same cycle. All these comple­mentizers may feature in noun phrase complements.

Owing to the fact that ku as well as kuti is a complementizer

formative, there is need to restate how the choice between the concords ku of class 15 and zvi of class 8 is determined. This could be done in the following way: with ku as complementizer formative there is a rela­tively free choice between the concords of class 8 and class 1 5 , except that, if the complement clause has -ti as its verb, the concord employed is restricted to that of class 8. However there is a tendency on the

whole to prefer the concord of class 15 to that of claas 8. With the other complementizer formatives only the concord of class 8 may be selected.

2*11 kuti /+" subjunctive/ and ku /+" infinitive/

The setting up of kuti subjunctive/ as well as ku /+ infinV/ as complementizers appears to help to account for yet another intriguing phenomenon with respect to the kuti clauses. For instance, consider thefoliowing pairs of sentences:

, „ x / / i t t /(161)a. Koni aida kuti adzoke.(Koni wanted to come back.)

/ / ✓ ib. Koni aida kuazoka.(Koni wanted to come back.)

/ / t t / / /(162)a* Koni aida kuti Farai adzoke.(Koni wanted Farai to come back.)

2 / / ibt, *Koni aida Farai kudzoka.(Koni wanted Farai to come back.)

The two sentences in (161) share the same meaning and they seem to be related structurally also. That is, it would appear that (161b) is

derivable from the same structure which • ' underlies (l6la). But in the

102

other pair only (162a) is well-formed, the other is not, yet It is con­structed on the pattern as (161b), How is this discrepancy to be accoun­ted for? Taking the underlying structure of (161a) and (l6lb) to be as

in (163)

(163)

Koni

("kuti subj j )V.k u Z F i n f i n T / / +N

+c l8+pro

Koni

dzoka

it would appear that (l6la) is derived by inserting kuti £+* subjunctive/ into S^* This yields the structure in (164)•

NPVKoni

da

h*N+CL8+pro

VR

Koni

dzoka

103

Applying the other rules which are not relevant to the discussion at hand we end up with (161a).

If to the structure in (163) we choose to insert ku Zt infin*7 instead, we get the structure in (165) "below®

(165)

VPNP.

V

S,da

+N+CL8+pro

ku*gL / T . NP, VP

Koni

dzoka

Fnom this structure (l6lh) is finally derived* Notice that in both (16^) and (165) Equi-NP deletion has to apply.

The underlying structure of (162a) and (162b) is much similar

to that of (161). It is given in (166) below.

(166)

Koni

* ua —I in[kuti £ f s u b 3 ^ 7 I [ku £ £ infinT7 /pj, —

W +GL8+pro

NP.

V

VP

Farai ' Vdzoka

I

104(167)

NP, VP

v- NP

S

+N+ CL8 *pro

kuti subQ_T7 ' NP,

.Farai y

dzoka

(168)

Koni

ku [ ± infin^7 ' NP.

Farai V

dzoka

105

Inserting kuti / +~ subjunctive/ as before into the result is the structure in (167)* From this tree is eventually obtained (162a). Notice that there is no Equi-NP deletion applying in this case. Now if instead

of kuti Z F subjunctivI7 we insert ku / ± infinitive/ into ^ in (166) the configuration we get is (168). From this derived phrase marker is even­

tually obtained the ungrammatical sentence (162b). What all this means is that ku £+ infinitive/ cannot be inserted into a structure like that in (166). But the question is: why does this transformation block in this case? This cannot be attributed to the verb in the higher S because

in (165) the same verb -da with the same meaning permits ku / T infinitive/ to be inserted into the complement clause. Notice that apart from the fact that the subject noun phrase of in (163) is Koni and that of in (166) is Farai these two tree diagrams are identical. In (165)

is deleted under identity with NP^, but in (168) this is not possible.So the ungrammaticality of (162b) appears to stem from the fact that there is an identity constraint which accompanies ku /+*" infinitive/ insertion. Perhaps this may be stated as a condition on Equi-NP deletion, but at the moment I do not know how that condition could be formulated.The subject noun phrase of the complement clause must be identical to either the subject noun phrase or the object noun phrase of the contain­ing S. The sentences in (169) and (170) are further illustrative examples of the type of sentences in (161) and (162) respectively.

(I69)a.i. Mwana^ abvuma kuti asare here?. . « / A- / , 1 f ■,1 1 . Mwana abvuma kusara here?

(Has the child agreed to remain behind?)f / / / / / / // /b.i, Ndakaedza kuti ndimubatsire asi zvakakona.

, / / / / / / / / / /1 1 . Ndakaedza kumubatsira asi zvakakona.

(I tried to help him but it was all in vain.)/ f / / / / / / / / ✓ / /c.i. Mbombo akaedzesera kuti arove nyoka ikabva yamusvipira.

, / / / / / / / / / / / /ii. Mbombo akaedzesera kurova nyoka ikabva yamusvipira.

106

(As Mbombo tried, to hit the snake it spate at him*)/ / t / /d*i, Muri kufungidzira kuti mudzoke rinhi?/ t *ii, Muri kufungidzira kudzoka rinhi?

(When are you expecting to come back?)

e.i, Mudzxdzxsi akanganwa kuti aridse dare./ / / /ii, Mudzidzisi akanganwa kuridza dare.

(The teacher forgot to ring the bell.)/ / / / / / /f.i. Vasekuru vabvumidza Kembo kuti aende kumba.f i / / tii* Vasekuru vabvumidza Kembo kuenda kumba.

(Grandfather allowed Kembo to go home.)(170)a*i* Wanatsa kuti vana/ vaende/ kuchiri* kuchena.

xxx (You did well to allow the children to go when it wasstill daylight.)

/ / i / fi fii.*Wanatsa vana kuenda kuchiri kuchena./ t / A / / / / / /b*i. Tomu akaedza kutx motokari yake ifambe.

(Tom tried to make his car move.)t i f f t t t i fii. *Tomu akaedza motokari yake kufaraba.

e.i, Anyangadza kuti vatezvara vagare pasil(He made a mistake in letting his father~in-law sit on the ground.)/ t /ii. *Anyangadza vatezvara kugara pasx,

/ / / / t i t / td.i. Sani akaramba kuti amai vake vatambudzike,(Sani could not tolerate seeing his mother suffer.)

/ / / / / t i txi. *Sani akaramba amai vake kutambudzika,e.i* Suzx abvuma kutx Masango amutakure nemota yake.

(Guzi agreed that Masango should give her a lift in his car.)

. . * „ t t, / „ / 1 , L f, * t * * . /, /i x . ^Suzx abvuma Masango kumutakura nemota yake*

In the case of (l?0e) if the complement clause Masango amutakure nemota / /

Yake is passivized we get (171).

10?

I t i t % / t i l t i t(171) Suzi abvuma kuti atakurwe naMasango nemota yake*(Suzi agreed to be given a lift by Masango in his car*)

Nov/ that the grammatical subject of this complement clause is identical

with that of the containing S, ku / + infinitive? insertion can apply without causing any difficulty* The resulting sentence is (172) belowwhich is perfectly well-formed*

t i t t i l t i t(172) Suzi abvuma kutakurwa naMasango nemota yake.(Suzi agreed to be given a lift by Masango in his car.)

2*12 kuti /T~ indicative? and ku /+ infinitive?The discussion immediately above involved the insertion of

kuti £+ subjunctive? and ku U infinitive? into the complement clause and the relationship, both structural and semantic, of the sentences: which result. A similar comparison obtains between kuti ?+ indicative? and ku /+ infinitive?. To this end consider the following sets of

sentences./ / / t 1 f t(173)a. Mwoyo akaona kuti akatadza.

(Mwoyo realized that he made a mistake.)b. Mwoyo akaona kutadza kwake^.

(Mwoyo saw his mistake.)/ / / t fc* *Mwoyo akaona kutadza.

(*Mwoyo saw erring.)/ / / / / / i t t t s ,(17A)a. Zikomu akafembera kuti Mushonga anosvika nhasi, iye achibva

i t Iasvikawo nhasi,(Zikomu guessed that Mushonga would arrive today, and for sure he arrived today.)S / / / / / t £ / / / s {b* Zikomu akafembera kusvika kwaMushonga nhasi, iye achibva

asvikawo nhasi.(Zikomu foretold Mushonga's arrival today, and indeed hedid arrive today.)

t / 1 s t t 1 / < ( tc* *Zikomu akafembera Mushonga kusvika nhasi, iye achibva

108f . ' i .asvikawo nhasi.(*Zikomu foretold Mushonga to arrive today, and for certain

he arrived today*)The (a) and (b) sentences share the same meaning in each set* Observe

that (173c) and (174c) which are identical to (173b) and (174b) except in one respect only, namely, the absence of the possessive kwake and kwaMushonga respectively, are ill-formed. That the (a) and (b) sentences in each set are related somehow structurally as well as semantically

seems obvious enough, and a linguistic description of Shona which does not capture this relatedness would be, to my mind, descriptively inade­quate.

What is interesting to observe here is that when ku / + infin77is inserted into complement structures governed by verbs like -ona and

/ / /-fembera the subject of the complement clause is not deleted by Equi-NP as was the case in the subsection immediately above, but is turned into a possessive construction. In this case there is no identity constraint

in respect of the insertion of ku / + infinitive? as was the case in the

preceding subsection, hence (173b) is well-formed and so is (174b). Here are further examples of the sentences in which ku Z±~infinitive/ inser­tion is accompanied by possessivization.

/ / / / / /Midzi ari kutsutsumwa kuti :(Midzi is unhappy because this affair became known.)

/ / f / / / / / f /Midzi ari kutsutsumwa kubuda pachena kwenyaya iyi.(Midzi is unhappy because this affair became known.)

/ / / / / / / / /*Midzi ari kutsutsumwa nyaya iyi kubuda pachena.(*Midzi is unhappy this affair to come into the open.)

/ /Ndaudza Nzira kuti zvinhu zvaipa.(I told Nzira that things were bad.)

/ / / /Ndaudza Nzira kuipa kwezvinhu*

(I told Nzira that things were bad.)

/ / / / / / / ✓ / / /(173)si. Midzi ari kutsutsumwa kuti nyaya iyi yakabuda pachena,

109e / /c* ^Ndaudza Nzira zvinhu kuipa.

(*I told Nzira things to be bad,)/ / / / / / i /.(177) a, Totama aivanza kuti ane mwoyo wakaipa./ / / / / / / / $b. Totama aivanza kuve kwake nemwoyo wakaipa,

(Totama used to hide his cruelty,)t / / / / i /c:» *Totama aivanza kuve nemwoyo wakaipa,

(*Totama used to hide cruelty,)

In some cases however possessivization accompanying ku /+ infinitive/insertion seems to be optional as in (178) - (180),

/ / / / / / / / / / / ✓ / / f(178)a, Mwana anotondera kuti akanzi naamai vake auye mangwana,b. Mwana anotondera kunzi kwake naamai vake auye mangwana.

f I t f / / / / / / / / /c. Mwana anotondera kunzi naamai vake auye mangwana,

(The child remembers having been told by .i;£s mother to come tomorrow,)

(179)a, Akarohwa akatsiidza kuti haachazviita zvakare,/ / / / / / /b. Akarohwa akatsiidza kusazviita kwake zvakare,

* s / Se„ Akarohwa akatsiidza kusazviita zvakare,(He was beaten so hard that he promised not to do it again.)

The problems which I leave as open questions are (a) how this possessive construction is to be characterized, and (b) when is this possessiviza- tion obligatory and when is it optional?

2,13 SummaryThe facts which have emerged so far in this chapter are (a)

that complementizers exist in Shona and these are listed in subsection2,10 above, (b) that the ku in kuti clauses dominated by NP is not a noun prefix but a complementizer formative whose phonological form

happens to be identical to that of the noun prefix of the head noun of the NP in which it occurs, (c) that there exists a complementizer form­

ative kuti as. well as kuti an infinitive form of the verb -ti, and (d)

that complement clauses which are dominated by NP may have as their

110head noun a pronoun of class 8 or class 15 normally, (I say i. ■. "normally" because there is a departure from this in a few cases as will be seen in 6.1*5*)

Ill

3*0 INFINITIVAL AND DERIVED NOMINALSThere are various types of nominal expressions which are found,

in Shona, In this chapter I should like to consider two of themi which are of special significance to this study. Let us call them infinitival and derived nominals. To this end examine the following sentences:

(1)a. Madzitateguru edu aipfeka nhernbe^(Our forefathers used to wear skin aprons,)

/ / /b* Ndapa vachati mari,(I gave the bride and bridegroom some money,)

/ /c, Tendere akafa.(Tendere died,)

d,. Vanaf vanoenda kumusha mangwana,(The children will go home tomorrow,)

/V * t i / t(2)a, kupfeka nh.em.be kwamadzitateguru edu(the dressing in skin aprons of our forefathers)

b,. kupa kwangu vachati mari7"(my giving the bride and bridegroom some money)

i / /c, kufa kwaTendere(Tendere’s death)

/ / / / /d, kuenda kwevana kumusha mangwana(the children's departure for home tomorrow)

/ / / . / / /(3) a, zv.ipfeko zvenhembe zvamadzit ate guru edu(clothing - of skin aprons - of forefathers - our = the clothing which consisted of skin aprons which our

forefathers had)* i / / /b, chipo changu chemari kuvachati

(gift - my - of money - to the bride and bridegroom =

my gift to the bride and bridegroom which was in the formof money)

/ / /c, rufu rwaTendere)

112(Tendere1s death)

/ ✓ / / /d. rwendo rwekumusha rwamangwana rwavana(journey - of home - of tomorrow - of children = the children's journey home tomorrow)

Corresponding to the sentences in (1) are the infinitival nominals in(2) and the derived nominals in (3)* There are many differences which have been noted between these two types of nominals in Shona linguistic studies. The differences that would strike one's attention most are those that pertain to the productivity of the transformation in question, the generality of the relationship between the nominal and the associa­ted underlying structure, and the morphological structure of the nominal.

The constructions which are being discussed here have their approximate counterparts in English in the form of gerundives and derived nominals. But since infinitival nominals are not always equivalent to the gerundives in English this term has ' ' accordingly been avoided

as it might be misleading in Shona, Chomsky discussed these gerundives]and derived nominals at some length, ' In his discussion of these two

types of nominal phrases he considers two positions, namely, what he labelled the "lexicalist position" and the "transformationalist position". In his conclusion he adopts the lexicalist position for the derived • nominals and the transformationalist position for the gerundive nominals.

Chomsky's conclusions here appear to be equally applicable to Shona in respect of the nominals which are under discussion. It seems to be the case that we shall achieve a large measure of descriptive adequacy if we adopt the transformationalist position for infinitival nominals and the lexicalist position for the derived nominals.

1. N*Chomsky in "Remarks on Nominalization" in Readings in English Transformational Grammar» Roderick A, Jacobs and Peter S0 Rosenbaum

(editors). (1970)

1133.1 Productivity

Consider first the question, of productivity# One thing which

we notice clearly is that we can form infinitival nominals relatively freely from underlying structures of the form noun phrase’ plus verb phrase. In fact there are very few exceptions indeed to this transform­ation., An example of such exceptions is a proposition in which the verb

/ / /- fnnira (ought to). However as far as the formation of derived nomi­nals is concerned there are many restrictions* The structures which underlie the sentences in(4) j for instance, are easily transformed intoinfinitival nominals of (3)j but not into the derived nominals of (6).

t / t f i t(A)a.. Murume uye akatarisa rutivi.(That man looked sideways.)

/ / / .h*. Musikana anyora tsamba*(The girl wrote a letter.)

/ / /c, hdege inobhururuka mudenga.(An aeroplane flies in the sky.)

/ / / /(3)a* kutarisa rutivi kwemurume uye(that man’s looking sideways)

/ / / / / /b*. kunyora tsamba kwemusikana(the girl’s writing of the letter)

c*. kubhururuka mude^nga kwendegey

(the aeroplanes flying in the sky)j t i i

‘ i t !*chitariso (7) rutivi chemurume uyeVs------ — ~ (other classes)^ / / / / / /b*. / ^munyoro (3) tsamba yemusrkana3 f / v a / / /\ *nyoro (3) tsamba remusikana

(other noun classes)c* ( *mabhururuko (6) mudenga endege

\ . . i /J *chibhururuko (7) mudenga chendege

vj;---------(other noun classes)

llifWhereas the structures which underlie hoth the sentences in (1) and those in (h) can have the transformational rules which yield infinitival nomi- nals applied to them, only those in (1) can he transformed into derived nominals. It seems quite clear that the transformation which gives rise to infinitival nominals is much more productive than that which would yield derived nominals were we to allow such a transformation in respect of derived nominals. There does not appear to he any rule hy which we

can predict the grammaticality of the phrases in (3 ) and the ungrammati- cality of those in (6)* Even where a proposition can be transformed into a derived nominal, on the whole there does not seem to be a way of pre­dicting which noun class, feature the derived nominal is going to have.The derived nominals in (3) clearly illustrate this point. The nominal

/zvipfeko in (3a) has class 8 feature but the nominal rufu in (3c) has a class 11 feature. If we assign a class. 11 feature to the proposition underlying (3a) we come up with the ungrammatical nominal phrase in (?}.

(7) *rupfeko rwenhembe^ rwemadzitateguru e duSimilarly if we assign a noun class 8 feature to the proposition which underlies (3c) or (3 -) the result are the ill-formed phrases in (8).

(8)a. *zvifu zvaTendere/ / / / /b. *zvendo zvekumusha zvamangwana zvevana

There are nevertheless a few cases where a noun class feature can be assigned to a derived nominal in a way which is more or less predictable. Such are the agentive derived nominals in (9) and the

abstract derived nominals in (10).(9)a.i. murimi (1) (a farmer)

ii. varimi (2) (farmers)/ /

b*ie muvhimi (1) (a hunter)/ /ii. vavhimi (2) (hunters)

e.i. mutongi (1) (a judge)/ /ii. vatongi (2) (judges)

cp -rima (plough)

cp -vhima (hunt)

cp -tonga (judge, try)

113

d.i* raudzidzisi (1) (a teacher)/ /ii. vadzidzisi (2) (teachers)

e.i. murwi (l) (a soldier)ii. varv/i (2) (soldiers)

(10)a. uroyi (14) (witchcraft, wizardry)

b. ugare (14) (peacefulness)c. urw/re (14) (an illness)

/ d. upfumi (14) (richness)e. uipi"(14) (evil, badness)

f* ufuro (1 4 ) (grazing area)

g. ungwaru (14) (cleverness)

s / . «cp -dzidzisa (teach)

cp -rwa (fight)

cp -roya (bewitch) cp -gara (sit, settle)

cp -rwara (be ill) cp -p-fumsT (be rich) cp -ipa^ (be bad)

cp -fura (graze)

cp -ngwara (be clever) cp -rema (be heavy)h» uremu (1 4 ) (weight)

But even then one does not go very far down any given list of verbs before one comes across counter-examples. The examples in (11) and (12)

counter those in (9) and (10) respectively.(11)a.i. *muendi (1) (a goer)

ii* *vaendi (2) (goers) b*i. *mumuki (1) (one who wakes up)

ii, *vamuki (2) (those who wake up) e.i, *muremi (1) (*one who weighs)

cp -enda (go)

ii. *varemi (2) (*those who weigh)/ /d*i* *mukwani (1) (*one who is enough)t * , , ,ii, *vakwani (2) (*those who are

enough)(I2)a, *uendi (1 4)

*uendo (1 4)*uanda (1 4)*uendu (14)*uende (14)

cp -muka (wake up)

cp -rema (be heavy)

cp -kwana (be enough)

cp -enda (go)

1X6b® *utarisa (14)

*utarise (IX)cp -tarisa (see, look)

*utariso (14)*utarisu (1 4 )

/ ✓ / /Co. *umiri (14) cp -mira (stand, stop)

Furthermore in (10), for instance, the terminal vowel may be /a/ as in, * , , s / . , , ,uroora, /e/ as in ugare, /i/ as in upfumi, /o/ as in ufuro, or /u/ as

in ungwaru» There does not seem to be a way of predicting in these nomi­nals what the terminal vowel is going to be® If in (10a) the terminal

t /vowel /a/ is altered to /e/ or /o/ or /u/ we get the forms furoore,// / /*urooro and *urooru which are all ill-formed* In (12c), on the other

hand, If the terminal vowel is changed to /o/ v/e get the well-formedt tderived nominal umiro (height; size)* It could be argued that derived

nominals of class 14 are themselves derivatives from derived nominals of other classes, notably classes 1 and 2* But the existence of such

counter-examples as the following appears to weaken the argument*

(1 3)a* *uvengi (14) cp muvengi (1) (enemy)b® *ubi (14) cp mubi* (1) (one who steals)c® *udzxdzi (14) cp vadzidzi (2) (learners)

It is clear that if these derived nominals are going to be obtained

transformationally a very complicated transformational apparatus will be required and it is doubtful whether this can be achieved without mak­ing the grammar look cumbersome and unwieldy*

3*2 MeaningThe second major difference between these two kinds of nominals

relates to the semantically idiosyncratic nature of the relationship bet­ween the derived nominal and the associated verb* Often derived nominals bear individual shades of meaning which sometimes differ considerably from the verb from which they are said to derive® The following illustra­tive examples will bear this out®

(14)a.i* muroora (1) (a daughter-in-law) ^ /

ii* roora (5) (bride price)iii. chiroora (7) (behaviour of a

daughter-in-law)

117cp -roora (marry

(with a male as subject in an act­

ive sentence))/• /b*i* demo (5) (an axe)

ii. matemo (6) (axes) cp -tema^(cut)

iii, nhemo (9/10) (a chisel/chisels)/ /e.i* mustingo (3) (a string)

ii. chisungo' (7) (a bond, a resolution, cp -sunga (tie)

d*i, garo (5) (a base; butt)ii* mugari (1) (a resident)iii. chigaro (7) (a chair) cp -gara (sit;iv* rugare (11) (peace) settle)v. ugare^(I k ) (peacefulness)

Although there are a few subregularities, for example in (9) above, on the whole however the range of variation in meaning is typical of lexi­cal structure. If it is accepted that it is the grammatical relationships

in deep structure which determine the meaning of a sentence, it follows that in order to account, within the transformational approach, for the facts which have been mentioned above it would be necessary to resort to the device of specifying a range of meanings for the base form. It would have to be stated also that with certain semantic features a form

must nomijiaiise but with others it must not* The grammar that is suggest­ed by this artifice will be ad hoc*

3*3 Internal StructureAnother observation which strikes one is the difference bet­

ween derived, nominals and infinitival nominals with regard to their internal structure. A derived nominal may have as one of its features

a noun class affix of almost any class* Locative affixes are however

118excluded isia this case unless they are super^imposed on other affixes*,

Morphologically a typical noun in Shona consists of a class affix and a stem as is illustrated in (15) below.

(1 5)a*, mu-rume (1) (a man)b* mi-kuyu (4 ) (fig-trees)c. ma-kuyu (6) (figs)d. chi-ngwa (7) (bread)e. pa-nze^ (16) (outside (near by))

/f. ku-nze (1 7) (outside (in general))Some derived nominals have a morphological structure which cosists ofa class affix, a verb root and a terminal vowel, the latter being deter­mined to some extent by the class affix*. Such are the nominals in (16).

/ / / /(16)a, mu-rtadz-i' (1) (a sinner) cp -tadza (sin; fail)✓ ✓ . . . .bo ru-tadz-o (11) (a sin)/ / / /c. mu-sung-o (3) (a string) cp -sunga (tie)/ /d8 chi-sung-o (7) (a bond)

e. rw-end~o (11) (a journey) cp -enda (go)/ / . / / /f. va-roor-a (2) (daughters-in-law) cp -roora (marry)

There are subregularities that may be observed here as was mentionedearliero For instance, the affixes mu- and va- of noun classes 1 and 2 respectively normally require that the unextended verb should have the

terminal vowel /i/ as in (16a) and also in (9). The kind of derived nominals in (16) may not be expanded to include either object noun

phrases or adverbials although the verbs themselves may occur with one verbal extension or more as illustrated in (1 7)*

(17)a* mu-sung-w-a (1) (a captive) cp -sunga (tie)/ * / /b ft chi-zvar-ir-w-o (7) (since birth) cp -zvara (beget)

c. ma-bik-ir-o (6) (manner of cooking) cp -bika (cook)

d0 zvi-sak-is-o (8) (causes) cp -saka (cause)

It is however also possible to have a derived nominal whose - morpho-

119logical structure comprises a class affix and a verb phrase* In this case the terminal vowel is almost I n v a r i a b l y limited to /a/* These nomi­nals seem to share the meaning of relative clauses* Here are a few exa­

mples of such nominals./ / /(l8)a. mu-tamb-a nengwe (la) (one who plays with a leopard)

cp -tamba nengw/ (play with a leopard)/ / ib 0 mu-komb-a choto (la) (one fond of warming oneself by the

fireplace)cp -komba chot/ (encircle a fireplace)

/ / ' * , ,c. chi-nzveng-a mutsvairo (7) (one who godges the broom -a lazy person)

/ / / /cp -nzvenga mutsvairo (dodge the broom)

d. chi-muts-a mapfihwa ( ?) (one who revives the hearth-stones= wife given to widower to replace the one who has died)

scp -mutsa mapfihwa (raise the hearth-stones)In (l$b - d) the verb phrases consist each of a verb and a direct object

noun phrase*The infinitival nominals resemble in their morphological form

the nominals in (18)* That is5 they always consist of the ku formative

and a verb phrase as shown in the underlined strings in (19)*(19)a. Mufudzi anogona kuti mombe dzluye^ masikati* (see (h9c.i) ir

Chapter 2)b. Mako anoda kufamba nevasikana. (see (51a) in Chapter 2)

This ku formative however as was seen earlier is not a class affix as such but a complementizer formative. Although the nominals in (18) and(19) both contain verb phrasess they nevertheless differ remarkably in other respects. The infinitival nominals are capable of much greater

complexity than the derived nominals* This complexity is in the direct­

ion of that which is found in sentences* Thus the following transform­ations^ which normally operate in sentencess may apply readily within infinitival nominals*

120(20)a* anaphoric pronominalization;

e.g. i. Jojo anoda kutenga mabhuku aya*(George wants to buy these books*)/ / / / , / ✓ , / vii* Jojo anoda kuatenga (namely f ::k •; mabhuku ay a)

(George wants to buy them (namelys these books))b. reflexive pronominalization;

/ / / / .// / / / / / /e*g* l, *Mbare akaramba kupxra Mbare kuvavengi vake*(*Mbare refused to give up Mbare to his enemies*)

ii* Mbare akaramba kuzvipira^kuvavengi!^ yak/.

(Mbare refused to give himself up to his. enemies.) c6 negation;

/ / /e.g* i, Mbare abvuma kuendako*

(Mbare has agreed to go there*) ii. Mbare abvuma kusaendako*

(Mbare has agreed not to go there*)d. pseudo-cleaving;

e.g. Mabhuku aya Jojo ndxwo aanoda kutenga* (cp (20a*i)

(These books are the ones which Goerge wants to buy.)e* interrogative pro-form;

/ / / / / * /e.g. Jojo anoda kutenga chii? (reply; mabhuku aya)(What does George want to buy?) (reply: these books)

Furthermore these infinitival nominals can contain aspect as in (21).V / / / / / / / / / /(21) Apedza kukanga nyama Muchaneta akatanga kuchibika sadza.

(When Muchaneta had finished frying the meat she began to cooksadza*)

An infinitival nominal which is in a subject position in a r sentence can still undergo any of the transformations listed in (20) as the sentences in (22) show*

(22)a* anaphoric pronominalization;/ / /ecg* Kumuona zvinonetsa.

121(It is difficult to see him.)

b. reflexive pronominalization;e.g. Kuzvi'sungirira zvakaipal

(Hanging oneself is bad.)

c. negation;/ / ? / / / /e.g. Kusabvakachirana nebama hazvina kunaka.

(It is not a good thing to keep away from one’s

relations*)d. pseudo-cleaving;

/ ✓ / , / /e.g. Jerusarema ndiro rinonakidza kutamba.(It is the Jerusarema dance which is interesting.)

cp„ Kutamba Jerusarema kunonakidza.

(The Jerusarema dance is interesting.)Anaphoric pronominalization has been attested in derived nomi­

nals whose morphological composition is as follows:class affix + verb + applied extension^ + /o/.

This is exemplified in (23) below.(23)a.i. Mabairo acho mombe iyi ane n g o z l 0

(Slaughtering this cow is dangerous.)/ / / / / .ii. Maibairo acho ane ngozx*

(.Slaughtering it is dangerous.)/ / / / ✓ / / /b.i. Maonero acho musikana iyeye anonetsa.

(Seeing that girl is difficult.)/ / / / / /xx0 Mamuonero acho anonetsa.

2. For the approach adopted in this study in respect of the Applied

extension see 6.1.A and also Appendix A. For other treatments of this extension see N.C.Dembetembe in Verbal Constructions in Korekore, an M.Phil. Dissertation5 London, (1969); also G.Fortune in Shona Grammatical

Constructions, Part I , (forthcoming).

122(Seeing her is difficult.)

/ / / / / / s s se.i. Matakuriro acho mango dzacho ndiwo andisingazive(How to carry the mangoes is what I do not know.)

/ / / * / , ,/ , / / ./ii* Madzitakurxro acho ndxv/o andxsxngazxve,(How to carry them is what I do not know.)

/ , / / / /

The class affix of the derived nominal in this case appears to he re-

stricted to that of class 6. There exists also an isolated example of a derived nominal with a negative formative.

Apart from these examples which are rather rare we do not on

the whole observe the transformations in (20) with derived nominals. These transformations block if applied to the structures underlying the nominals in (16) through (18). In the following sentence the nominal in(l$b) appears as a direct object noun phrase,

/ / / / /(2k) Garwe aona mukomba-choto»

If, for instance, we apply anaphoric pronominalization, the negative and the pseudo-cleaving transformations to the structure underlying (2k)j we eventually get the following sentences in (25) which are however

all ungrammatical.

(25)&* anaphoric pronominalization;( /*Garwe aona muchikomba.

(Garwe has seen the one who is fond of sitting by it.)b* negation;

f * , / , ' , '*Garwe aona musakomba choto.(Garwe has seen the one who is not fond of sitting by the fireplace,)

c, pseudo-cleaving;

(mischief; disobedience) cp -nzwa (obey)

(Garwe has seen the one who is fond of sitting by the fire­

place. )

123^ t / / / /*Garwe, choto ndieho chaaona mukoraba*

(*The fireplace is what Garwe has seen the one who encircles) Let us consider further the following sentences:

(2 6 )a. Taka haadi^ rova-mb'ira.(Taka does not want a rock-mamba*)

b* Musikana anofarira muda-vanhu*(The girl is happy to see the one who loves people*)

/ / / / , /Co Tmotanga munzvenga-dare»(We start with the one who stays away from the dare*^)

(27)a* Taka haadi''kurova mbira*(Taka does not want to kill rock rabbits*)

* / / / s f . / / .b. Musikana anofarxra kuda vanhu*

(The girl enjoys loving people*)/ / , / / /c, Txnotanga kunzvenga dare,

(We start by avoiding the dare*)✓ y y / / /(28)a* *Mbira hadzidi rohwa naTaka*

(*Rock rabbits do not want be killed by Taka*) b e *Vanhu vanofarxra mudxwa nemusikana*

(^People are happy the one who is loved by the girl*)✓ . / / / / /c* *Dare rxnotanga munzvengwa nesu*

(*Dare begins the one avoided by us*)✓ / / / / /(29)a* Mbira hadzidi kurohwa naTaka.

(Rock rabbits do not want to be killed by Taka.)/ / / / / / / / /b* Vanhu vanofarira kudlwa nemusikana.

(People like to be loved by the girl*)

c* Dare rinotanga kunzvengwa nesu*(The dare will be avoided by us first.)

5® A dare is a meeting place for men living in the same village. They have their meals there.

12kThe strings which are underlined in (26) and (2?) are derived nominals and infinitival nominals respectively* Both types of nominal are funct­ioning as direct object noun, phrases* All these nominals contain verb phrases each of which consists of a verb and a direct object noun phrase But notice that whereas the infinitival nominals in (27) can undergo the passive transformation as shown in (29)> the derived nominals in (2 6)

cannot as (28) shows. The derived nominals in (26) form tight construct­ions whose morphology afed word order cannot be disturbed without destroy ing the nominal. Infinitival nominals on the other hand are flexible*Some of their constituents can move around outside the nominal itself.

/ / fFor instance? in (29a) mbira has moved outside the infinitive kurohwa.In fact in (29a) the infinitive nominal is now kurohwa*' naTakal

lar in their linear structure, their derivation is remarkably different* Assuming for argument’s sake that these two sentences have a similar underlying structure their deep structure representation would be as in (30) and (3 1) respectively*

Although, for example, (26b) and (2?b) look superficially simi

(30)

NP.1 VP

musikana V NP

farira

da vanhu

I 125

(51)

musikana NP

farira 1

VPNP+N+CL8+pro

Vmusikana

da

h

va'nhu

In (31) after ku £+ infinitive/ has been inserted into 5 NP^ can bedeleted under identity with NP^* Then applying the other rules whichare not essential to this discussion the result will be the sentencein (27b)* But in (30) there is a problem in trying to derive, the nominal

/ /muda-vanhu transformationally* To begin with our intuition tells us/ /that the subject noun phrase of cannot be musikana because she is

not the one who loves the people* She is only being charmed by the onewho loves the people* What is clear also is that what is loving thepeople is some human being* In order to accommodate_thts subject I have

/+ pro,/put down the subject noun phrase o£T with A p ers/ feature fill in this gap* The question for which a satisfactory answer is not immedia­tely forthcoming is by what rule is this subject NP deleted as beco­mes nominalized? We cannot invoke Equi-NP deletion because its struct­ural description will not have been met, that is, NP^ is not Identical

with NP <> If this approach is maintained NP^ will have to be deleted in some ad hoc fashion and this obviously is unsatisfactory* Faced with this and other related problems which we saw above it is difficult to see how a transformational approach to derived nominals can be use­fully adopted*

126In this study the lexicalist approach will be adopted for

derived nominals and the transformationalist approach for infinitival nominals*. A derived nominal like ungwaru (see (10g)) would include the following among its feature specifications in the lexicon:

(32) /ngwara, +N, +CLlif, +u — — J

The word ngwara will then be spelt out phonologically as ungwaru in this case* What (32) means id that the whole thing within the square brackets is a nominal because of the presence in the feature specifica- tion of the element £+If_/* Ibis nominal will have a noun affix of class lif and^ final vowel /a/ of ngwara will be replaced by the vowel /u/*Of course the feature specifications in (32) will have to include also the meaning of ngwara* Since derived nominals are not going to play a

significant role in the rest of this study thalgr discussion will not be pursued any further. Our concern will be with infinitival nominals and the other constructions in which complementizei's are found*

127

if.O COMPLEMENT!ZERS AND THE UNDERLYING STRUCTUREif#.l The Trans formational Hypothesis

In my discussion so far I have assumed that complementizers

are introduced transformationally into embedded complement sentences.'1*

Following Bresnan I shall refer to this, approach as the transformation-2alist hypothesis* For the present I would like to discuss the extent

to which this assumption can be validated.

It has already been noted that, although there are no identity constraints on the NP in the complement clauses, there are nevertheless restrictions that are imposed on the nature of a complement clause by

the verb of the higher sentence. For instance, in (1) below the comple­ment sentence is declarative but in (2) the complement sentence is a question*

✓ / / /(1) Ruzvidzo aida kuba mango.(Ruzvidzo wanted to steal some mangoes,)

(2) Sara abvunza kana Suzi^akauya.(Sara asked whether Suzi had come,)

This is illustrated by the tree diagrams in (3) and (h) which represent roughly the structures which underlie (1) and (2) respectively.The complementizer which is introduced into the embedded sentence in (3) is ku /+ infinitive^/, but in. (k) it is kana [ + participial^/. (The list of complementizers in Shona is given in 2,10.)

1* For grammatical analyses which take this view see G.Lakoff (196®);

Peter S.Rosenbaum (1967); J.R.Ross (1967); R.Lakoff (1968).2. Joan W.Bresnan in ”0n Complementizers: Toward a Syntactic Theory of

Complement Types” in Foundations of Language, 6 (1970), p.297 - 321®

da_£hku +/infin77 \imam «*«• 1

+ GL15+pro

Ruzvidzo V

mango

(k)

NP VP

Sara

_ bvunza^■fkana /-i-partic *JJ

N S1

~N “+CL8

PreS VP

Suzi

uya

Tree diagrams (3a) and (ka) show ku /+ infinitive^/ and kana /+ partic_*7 inserted into (3) and ( k ) respectively.

129

(3)a

Ruzvidzo V

da

NP VPku _ /+infin/

Ruzvidzo NP

ba mango

(if)a

Sara

bvunza

VPNP4-N +CL8 + pro

Suzi

uya

Note that when an embedded question is introduced by kana /+ participial/ the question formative does not appear in an overt form as is further

tillustrated by (kh) and (he) below*

t / / r p i / / /(k)h* Sara abvunza //Suzi akauya here2/l £ P j

NP NP(Sara asked, did Suzi come?)

/ / / / / /c, Sara abvunza kana Suzi akauya*(Sara asked whether Suzi came,)

130

/ a.Here in (4b) is question formative which does not appear in (4c).

The insertion of ku A infinitive^/ and kana /+ participial^/into (3a) and (4a) respectively is governed hy the verbs in the contain-

/ _m g sentences, namely, -da and -bvunza. If ku infinitive^/ is inser­

ted into the embedded sentence in (4)? and kana /+ participial^/ into the embedded sentence in (3)9 the result are the ungrammatical sentences in (5) and (6) respectively. The complementizer formatives are underlined in these examples.

(5) *Sara abvunza Suzi^kuuyaC(6) '^Ruzvidzo aida kana aibsf mango.

The point is that the verb ■ -da may not be followed by a complement clause which is introduced by kana £+ participial^/. The same applies to the verb -bvunza; ijs cannot be followed by a complement clause which

is introduced by ku /+ infinitive^/. Therefore it seems quite clear that the verb in the higher sentence is the one which determines which comple­mentizer is to be inserted into an embedded sentence.

It is obsei’ved that ku and kana which are present in the sentences in (1) and (2) respectively are not present in the deep stru­ctures of these sentences as represented by the trees in (3) and (4).The fact is, according to this approach, complementizers are regardedas mere markers of subordination which are introduced into embedded sentences indicating that these embedded sentences ■ were generated by the operation of the phrase structure rule:

NP ________^ N(S).

Furthermore, given "meaning preservation" of transformational rules,

this hypothesis claims that complementizers are devoid of semantic con­tent per se. They are not generated in deep structure, but are rather introduced transformationally.

Some of these complementizers are grammatical in some sentences but not in others. Which complementizer is found in any given embedded

131

sentence is governed by the verb in the higher sentence* It seems to be

the case also that most verbs can occur with only one complementizer*

Some however are capable of occurring with more than one though not in one and the same complement clause, for complementizers are mutually exclusive* The following are examples of verbs which may occur with more than one complementizer;

(7)a* Masanjara agona kuti aend/ mvura isati yanaya*

(Masanjara -- did well - that - he go ~ rain - it not having -

it rained ^ Masanjara did well to go before it rained*). f t / t t / / /b 0 Masanjara agona akaenda mvura isati yanaya.

(Masanjara did well to go before it rained*)In (7a) the verb -gona is accompanied by the complement clause kuti aende

which is introduced by kuti /+ subjunctive_7« But in (7b) the same verb -gona is followed by the complement clause akaenda which is introduced by kana /+ consecutive_7. These two sentences have the same deep struc­ture and the same semantic interpretation* (Note that kana is then obligatorily deleted in this sentence (see 2,8)) The verb -gona in thiscase is subcategorized as occurring in the environment /” ______ S 7. ThisS is not dominated by NP, but directly by VP. When the verb -gona is

subcategorized in this way, then either kuti /” subjunctive_7 or kana / + consecutive_7 must be selected and inserted into the complement clause. The deep structure of the sentences in (7) is provided in (8) below.

(8) S

Masanjara V

gonaC+kuti ,/+ subj.^7 ) y+kana /+ consec^ / \

JIDV7+ time 7

Masanjara V

enda " mvura isati yanaya

132

Here are.' further examples of pairs of sentences in which the individual sentences in each pair have different complementizers hut sharing the same meaning® The complement sentences are underlined and the compleme­ntizers are enclosed in square brackets after each sentence®

t t / / / _(9)a* Ngatitangei tadya ndokuzoenda kubhora* /kana + participial_/t i t / / _ _b® Ngatitangei kudya ndokuzoenda kubhora® /ku + infinitive^/

(Let us have our meal first and then go to watch soccer afterwards*)

t / / / ~(10)a® Dambudzo asara akasvika iwe waenda. /kana + consecutive^// / / / / / . —b® Dambudzo asara osvika iwe waenda® /kana + participial^/

(Dambudzo - remained - she arrive - you - having gone =

Dambudzo arrived after you had gone®)

(11)a. Nezuro vakarara vachitamba^ ngoma® /kana + participial^/ (yesterday - they slept - they dancing - a drum =They spent last night dancing to the tune of a drum®)

b* Nezuro vakararotamba ngoma®^ /ku + infinitive^/

(yesterday - they slept - dance - a drum =

They slept last night dancing to the tune of a drum®)(12)a® Sekai naJeni^vanogara vachitukana® /kana + participial^/

(Sekai - and Jeni - stay always - they scolding each other =Sekai and Jeni are always scolding each other*)

/ f t / / / _ _b® Sekai naJeni vanogarotukana* /ku + infinitive^/(Sekai - and Jeni - stay always - scold each other =Sekai and Jeni are always scolding each other*)

It is not uncommon to find a verb which takes any two or more complementizers usually with some difference in semantic interpretation®

The forms -tamba in (11b) and -tukana in (12b) are contracted

infinitives which are discussed in 6®2*1®2*

133

In such instances it seems to be the case that the deep structure in which that verb occurs with one complementizer will be different from that of the same verb with another complementizer * As an illustrationconsider the following pair of sentences,

i f f t(13)su Piri agona kuvhura gonhi®(Piri managed to open the door,)

b* Piri" a^gona akavhura gonhi*(Piri - did well - she opened - the door =Piri did well by opening the door,)

The complement clause in. (13a) is introduced by ku /+ infinitive^/ while

that of (13t) is introduced by kana [ + consecutive^/* The meaning of the verb -gona is different in these two sentences. In (13a) it means "manage, succeed", whereas in (13b) it means something like "be wise, clever, do well". According to this approach the difference between these two sentences is attributed, not to the different complementizers, but rather to the different semantic readings of the verb -gona. That there are two distinct verbs which happen to have the same phonological form -gona is further confirmed by the parallel differences of meaning with the following sentences some of which do not contain complementizers®

t v m . / / / /(in)at, Tsxtsx agona samu*(Tsitsi got the sums right*)

1 / / / ,b, Tsitsi agona akavhura gonhx.(Tsitsi did well by opening the door,)

, s t t t t(13) Tsitsi agona akagona samu,(Tsitsi did well by getting the sums right*)

With the meaning which it has in (l^a) -gona can occur with or without a complement clause, whereas -gona in (lAb) must always occur with a complement clause* In (13) the first -gona is the same as the one in (lAb), while the second -gona is the same as the one in (lAa)„ So they are definitely two different verbs which happen to share the same form*

13k

One: interesting observation is that verbs of "counselling" and of "asking" like: -udza (tell), -kumbira (ask), -kolrera (instruct forcefully), -raira (give advice, counsel), -tema (decide, decree), can­not normally have the subject noun phrases of the higher sentence and the lower sentence identical* The sentences in (16) are grammatical

but not those in (1?)*/ / / / / / / / / / s(16)a* Taka akakumibira Deni kuti abve mumba imomo*

(Taka asked Deni to move out of that house.)/ / y t s / / / / / /b*. Taka akamukumbira kuti abve rnumha imomo.

(Taka asked hdmi to move out of that house*)t i / / / / / / / / / n(17) a* *Taka akakumbira Taka kuti abve liiumba imomo *

(*Taka asked Taka to move out of that house*) b* *Taka/ aka z vikumbir a/ kuti abve mumba imomo*

(■*Taka asked himself to move out of that house.)But verbs whieh refer to mental state and also those which relate to speaking can have the subject noun phrases of the higher sentences and

lower sentences identical. These, are verbs like: -ti (say), -funga (think), -rangarira (remember, ponder, meditate), -yeuka (remember), -fungidzira (suspect, expect), -karira (expect, think), -ona (see, real­ize)*. In the deep structure of the following sentences the subject noun phrases of the complement sentences, are identical with those of the respective higher sentences* In the superficial structure however the subject noun phrase of the complement sentence does not appear because

it is deleted obligatorily by the ^qui-NP rule*, / i i /(18) Mwana afunga kuenda Joni*

(The young man has thought of going to Johannesburg*)/ l i i l / / /(19) Vasekuru vakati vanouya mangwana.

(Grandfather said that he would, come on the. following day.)

k* This sentence would be grammatical if the second Taka did not refer

to the first Taka*

135T h e r e f o r e i t w o u l d a p p e a r t h a t t h e r e a r e r e s t r i c t i o n s t h a t a r e f o u n d

b e t w e e n t h e s u b j e c t n o u n p h r a s e s o f h i g h e r a n d l o w e r s e n t e n c e s w i t h s o m e

c l a s s e s o f v e r b s „

C o n s i d e r f u r t h e r t h e s t r i n g s w h i c h a r e u n d e r l i n e d i n t h e f o l ­

l o w i n g p a i r o f s e n t e n c e s *

t t . * , / / / / /,( 2 0 ) a * N h a m o p a a p u n z i k a , k u t i a s i m u k e z v a r a m b a .

( N h a m o - w h e n h e f e l l - t h a t - h e s t a n d u p - i t r e f u s e d =

A f t e r N h a m o h a d f a l l e n d o w n h e f a i l e d t o g e t u p o n h i s f e e t . ^

/ / / / / / / / / /b . N h a m o p a a p u n z i k a , a r a m b a k u t i a s i m u k e *

( N h a m o - w h e n h e f e l l - h e r e f u s e d - t h a t - h e g e t u p =

W h e n N h a m o f e l l d o w n h e r e f u s e d t o g e t u p o n h i s f e e t * )

I n b o t h t h e s e s e n t e n c e s t h e v e r b i n t h e h i g h e r s e n t e n c e i s - r a m b a a n d

/ / /t h e c o m p l e m e n t c l a u s e i s k u t i a s i m u k e * B u t n o t e t h a t w h e r e a s i n ( 2 0 a )

t h i s c o m p l e m e n t c l a u s e i s i n a s u b j e c t p o s i t i o n , i n ( 2 0 b ) i t i s i n a

d i r e c t o b j e c t p o s i t i o n . B e s i d e s , t h e m e a n i n g o f t h e v e r b - r a m b a i s d i f ­

f e r e n t i n t h e t w o s e n t e n c e s . I n ( 2 0 a ) i t m e a n s " f a i l , b e u n a b l e " b u t i n

( 2 0 b ) i t m e a n s " r e f u s e , b e u n w i l l i n g " . I t m a y a l s o b e o b s e r v e d t h a t t h e

c o m p l e m e n t c l a u s e i n b o t h c a s e s i s i n t r o d u c e d b y t h e s a m e c o m p l e m e n t i z e r ,

n a m e l y , k u t i /+• s u b j u n c t i v e _ / 0 W h a t i s s t r i k i n g h e r e i s t h a t - r a m b a i s

i n t r a n s i t i v e i n ( 2 0 a ) b u t t r a n s i t i v e i n ( 2 0 b ) . W h e n i t i s i n t r a n s i t i v e

t h e s u b j e c t n o u n p h r a s e m u s t b e . a b s t r a c t i n t h e s e n s e t h a t i t m u s t b e

e x p a n d e d a s : N ( S ) , a t s o m e s t a g e . B u t w h e n i t i s t r a n s i t i v e i t s d i r e c t

o b j e c t m a y o r m a y n o t b e a n a b s t r a c t n o u n p h r a s e .

T h e s e o b s e r v a t i o n s s u g g e s t t h a t t h e s e n t e n c e s i n ( 1 3 ) s a n d

a l s o t h o s e i n ( 2 0 ) , d i f f e r f r o m e a c h o t h e r i n t h e i r d e e p s t r u c t u r e . T h e

v e r b - g o n a i n (13a ) a n d - g o n a i n (13b ) d i f f e r f r o m e a c h o t h e r i n b o t h

Rt h e i r m e a n i n g a n d t h e i r r u l e f e a t u r e s p e c i f i c a t i o n , e . g . t h e y h a v e

5 . F o r a d i s c u s s i o n o n r u l e f e a t u r e s s e e G . L a k o f f i n I r r e g u l a r i t y

I n S y n t a x , ( 1 9 7 0 ) .

1.36

d i f f e r e n t c o m p l e m e n t i z e r s * I n t h e c a s e o f ( 2 0 ) t h e t w o v e r b s w i t h t h e

s a m e p h o n o l o g i c a l f o r m - r a m b a d i f f e r i n t h e i r m e a n i n g a s w e l l a s i n t h e i r

s e l e c t i o n a l r e s t r i c t i o n s , . N o n e t h e l e s s t h e i r r u l e f e a t u r e s p e c i f i c a t i o n

w i t h r e s p e c t t o c o m p l e m e n t i z e r s i s t h e s a m e a s w a s p o i n t e d o u t a b o v e *

S i n c e , a c c o r d i n g t o t h i s a p p r o a c h , c o m p l e m e n t i z e r s a r e i n t r o ­

d u c e d i n t o e m b e d d e d s e n t e n c e s b y t h e o p e r a t i o n o f t r a n s f o r m a t i o n a l r u l e s ,

t h e v e r b s w h i c h a r e s u b c a t e g o r i s e d a s h a v i n g s e n t e n t i a l c o m p l e m e n t s w i l l

b e e n t e r e d i n t h e l e x i c o n t o g e t h e r w i t h t h e c o m p l e m e n t i z e r s w h i c h t h e y

p e r m i t t o b e i n s e r t e d i n t o t h e i r c o m p l e m e n t c l a u s e s * T h e c o m p l e m e n t i z e r s

w i l l f o r m p a r t o f t h e r u l e f e a t u r e s p e c i f i c a t i o n o f t h e s e v e r b s , a r u l e

f e a t u r e s p e c i f i c a t i o n b e i n g c o n s i d e r e d a s a n i n s t r u c t i o n i n d i c a t i n g

w h i c h r u l e o r r u l e s a p p l y t o t h e s t r u c t u r e i n q u e s t i o n * T h i s r u l e s p e c i ­

f i c a t i o n w i l l f o r m p a r t o f t h e c o m p l e x s y m b o l o f t h e v e r b o f t h e c o n t a i n ­

i n g s e n t e n c e * T h e s e r u l e f e a t u r e s i n d i c a t e w h i c h r u l e r e l a t i n g t o c o m p l e ­

m e n t i z e r s m a y b e i n s e r t e d i n t o a c o m p l e m e n t c l a u s e * T h e c o m p l e m e n t c l a u s e

m a y ' e i t h e r b e u n d e r t h e d i r e c t d o m i n a t i o n o f V P o r N P , a n d i n t h e c a s e

o f t h e l a t t e r t h e c o m p l e m e n t c l a u s e m a y b e i n e i t h e r t h e s u b j e c t o r t h e

o b j e c t p o s i t i o n i n r e l a t i o n t o t h e v e r b i n t h e c o n t a i n i n g s e n t e n c e * A f t e r

a c o m p l e m e n t i z e r h a s b e e n i n s e r t e d i n t o a n e m b e d d e d s e n t e n c e t h e s t r u c ­

t u r e w h i c h r e s u l t s i s s t i l l n o t t h e f i n a l f o r m o f t h a t s e n t e n c e w h i c h

i s u n d e r c o n s i d e r a t i o n * M o r e o f t e n t h a n n o t , o t h e r t r a n s f o r m a t i o n a l r u l e s ,

s o m e o f w h i c h a r e o b l i g a t o r y a n d o t h e r s o p t i o n a l , m u s t b e a p p l i e d t o

t h e s e c o m p l e m e n t - c o n t a i n i n g s e n t e n c e s b e f o r e t h e s u r f a c e s t r u c t u r e i s

r e a c h e d *

T o i l l u s t r a t e f u r t h e r t h e c o n c e p t o f r u l e f e a t u r e i n r e s p e c t

o f s e n t e n t i a l c o m p l e m e n t a t i o n i n S h o n a l e t u s l o o k a t t h e s e n t e n c e s i n

( 1 ) a n d ( 2 ) m o r e ' c l o s e l y * T h e c o m p l e m e n t i z e r s w h i c h w e h a v e s e e n i n t h i s

r e g a r d a r e : k u / + i n f i n i t i v e _ / a n d k a n a / + p a r t i c i p i a l ^ * I n t h a t c a s e

t h e v e r b s - d a a n d - b v u n z a w o u l d b e m a r k e d i n t h e l e x i c o n ( e i t h e r d i r e c t l y

o r b y a r e d u n d a n c y r u l e ) a s r e s p e c t i v e l y + ^ k u / + i n f i n i t i v e 7 7 *

137

•^ku 1+ infinitive // and -^kana £+ participial //, +</kana /+ particvTZ.

Since complementizer insertion is sensitive to the rule feature of the verb in the higher sentence? its operation is peculiar in that a complementizer cannot be inserted into a clause S^ during the trans­

formational cycle on that S^» A complementizer can only be inserted into this S^ during the cycle on the higher sentence which dominates

The tree diagrams in (21) below show how (1) is derived following this transformational hypothesis. Diagram (21a) is the same as the one in (j5) which is reproduced here for convenience.

( 21)a.

Ruzvidzo

Z+ku 1+ infinitT/ ^+N+CL15-i-pro

Ruzvidzo

mango

(21)b0

Ruzvidzo V

da '1

VP

Ruzvidzo NP

ba mango

Euzvidzo V NP

V P

NP

b a m a n g o

I t i s o b s e r v e d t h a t o n t h e f i r s t c y c l e i n ( 2 1 a ) n o t r a n s f o r m a t i o n s a p p l y

i n t h i s c a s e * ^ O n t h e s e c o n d c y c l e , t h a t i s , k u £ + i n f i n i t i v e / i n s e r -

✓t i o n a p p l i e s b e c a u s e t h e v e r b - d a i s m a r k e d f o r i t * T h i s i s i l l u s t r a t e d

i n ( 2 l b ) * I n ( 2 l c ) t h e s u b j e c t - N P o f h a s b e e n d e l e t e d u n d e r i d e n t i t y

w i t h t h e s u b j e c t - N P o f S Q *

I f w e h a d a s e n t e n c e l i k e t h e o n e i n ( 2 2 ) i t s d e r i v a t i o n a l

h i s t o r y w o u l d b e a s s h o w n i n ( 2 3 ) *

( 2 2 ) J e k e a i d a k u t i m o t a / t y a i r w e n a B e n i *

( J a c k w a n t e d t h e c a r t o b e d r i v e n b y B e n * )

B e c a u s e t h q r e i s a p a s s i v e m a r k e r i n i n ( 2 3 a ) t h e p a s s i v e t r a n s f o r m ­

a t i o n i s a p p l i e d o n T h e c o n f i g u r a t i o n w h i c h r e s u l t s i s ( 2 3 b ) e T h i s

c a n b e d o n e b e c a u s e t h e p a s s i v e t r a n s f o r m a t i o n a l r u l e i s n o t s e n s i t i v e

t o t h e v e r b i n t h e . h i g h e r S f o r i t s a p p l i c a t i o n * A f t e r t h i s a p p l i c a t i o n

6* For a discussion on the concept of transformational cycle refer toN cChomsky in Aspects of the Theory of Syntax (1963); also J 0T„Grinder

and S eH«Elgin in Guide to Transformational Grammar (1973); J*P«Kimball

on ’'Cyclic and Linear Grammars”, and John Grinder on "On the Cycle in Syntax" both in Syntax afad Semantics, (1972), J eP,Kimball (editor)*

I139

(23)a

Jeke

VPNP

passive Beni

tyaira raota

(23)bo

Jeke

VPNP

PPmo t a

P NP

I ina Beni

of the passive rule we cannot tell what rule to apply next on SinceS^ is- an embedded sentence, we look up to the verb in the higher S to see vrtiat rule feature(s) it has* In the case at hand our attention .c centres on the verb ~da6 Since it is marked for kuti /+ subj«_/

the next step is to insert kuti /T subjunctive^ into Sx as in (23c).

After applying later rules we finallyncome up with sentence (22)*

In both (21) and (23) the crucial point is that no compleme­ntizer insertion rule could apply on the first cycle, S1 , because the

140

(23)c*

N P V P

V N P

V PN P

m o t a P P

I / \t y a i r w a P N P

i , i .n a B e n i

t r a n s f o r m a t i o n w o u l d n o t ‘k n o w 1 w h i c h c o m p l e m e n t i z e r ( s ) i s p e r m i t t e d

b y t h e v e r b - d a * u n t i l t h e n e x t h i g h e r S ( i * e * S ) i s c y c l e d , , I n o t h e r

7w o r d s , a s B r e s n a n s a y s , t h e s t r u c t u r a l d e s c r i p t i o n o f a n y c o m p l e m e n t i z e r

i n s e r t i o n t r a n s f o r m a t i o n c a n n o t b e l i m i t e d t o a c o m p l e m e n t c l a u s e , b u t

m u s t i n c l u d e t h e v e r b v / h i c h t h a t c l a u s e c o m p l e m e n t s * T h i s t r a n s f o r m a t i o n ­

a l h y p o t h e s i s h a s . t h e a d v a n t a g e t h a t i t e n s u r e s t h a t n o n - e m b e d d e d s e n t ­

e n c e s c a n n e v e r a p p e a r w i t h c o m p l e m e n t i z e r s , e , g *

, . t / / / /(24) *Kuti mota ityairwe naBeni*( * T h a t t h e c a r b e d r i v e n b y B e n * )

4*2 Inadequacies of the Transformational HypothesisW h i l e i t i s t r u e t h a t t h e t r a n s f o r m a t i o n a l h y p o t h e s i s g u a r a n ­

t e e s t h a t n o n o n - e m b e d d e d s e n t e n c e s w i l l e v e r a p p e a r w i t h c o m p l e m e n t i z e r s ,

i t n e v e r t h e l e s s u n d e r m i n e s a n o t h e r w i s e w e l l - m o t i v a t e d u n i v e r s a l w h i c h

v/as p u t f o r w a r d b y C h o m s k y , n a m e l y , t h a t w h i l e t r a n s f o r m a t i o n s m a y r e -

7 * o p * c i t * p e 2 9 9

141

m o v e m a t e r i a l f r o m e m b e d d e d s e n t e n c e s , n o m o r p h o l o g i c a l m a t e r i a l c a n b e

i n t r o d u c e d i n t o a s t r u c t u r e w h i c h i s d o m i n a t e d b y S o n c e t h e c y c l e o f

t r a n s f o r m a t i o n a l r u l e s h a s a l r e a d y c o m p l e t e d i t s a p p l i c a t i o n t o t h i s

s t r u c t u r e . ^ F o r i n s t a n c e , a s w e s a w i n ( 2 1 a ) a n d ( 2 3 b ) c o m p l e m e n t i z e r

i i r s e r t i o n c o u l d n o t a p p l y w h e n w a s c y c l e d b e c a u s e a t t h a t s t a g e i t

w o u l d n o t b e k n o w n w h i c h c o m p l e m e n t i z e r w a s t h e a p p r o p r i a t e o n e t o b e

i n s e r t e d * C o m p l e m e n t i z e r s w e r e i n s e r t e d o n l y i n t h e n e x t h i g h e r c y c l e ,

n a m e l y , S Q i n b o t h c a s e s . C o m p l e m e n t i z e r i n s e r t i o n u s u a l l y i n t r o d u c e s

n e w m o r p h o l o g i c a l m a t e r i a l i n t o t h e S t o w h i c h i t a p p l i e s a s w a s s e e n

i n t h e e x a m i p l e s c i t e d a b o v e .

O n e c l a i m m a d e b y t h e t r a n s f o r m a t i o n a l h y p o t h e s i s i s t h a t

c o m p l e m e n t i z e r s a r e s e m a n t i c a l l y e m p t y . W h i l e t h i s m a y a p p e a r t o b e

t r u e w i t h r e s p e c t t o t h e p a i r s o f s e n t e n c e s i n ( 7 ) - ( 1 2 ) , i t i s c e r t ­

a i n l y n o t t r u e o f t h e p a i r o f s e n t e n c e s i n ( 2 3 ) ,

( 2 5 ) a , M i d z i a k a n y a r a k u t i a k a t a u r a n h e m a p a d a r e .

( M i d z i — w a s a s h a m e d — t h a t — h e t o l d — l i e s — a t c o u r t =

M i d z i w a s a s h a m e d o f h i m s e l f b e c a u s e h e l i e d a t t h e c o u r t , )

b , M i d z i a k a n y a r a k u t i a . t a u r e n h e m a i p a d a r e .

( M i d z i - w a s a s h a m e d - t h a t - h e t e l l - l i e s - a t c o u r t =

M i d z i w a s t o o a s h a m e d t o t e l l l i e s a t t h e c o u r t . )

T h e s e t w o s e n t e n c e s i n ( 2 5 ) a r e d i f f e r e n t i n m e a n i n g a l t h o u g h t h e y

w o u l d d e r i v e f r o m t h e s a m e u n d e r l y i n g s t r u c t u r e a c c o r d i n g t o t h e t r a n s ­

f o r m a t i o n a l h y p o t h e s i s . T h e i r u n d e r l y i n g s t r u c t u r e . i s p r o v i d e d i n (26a )

b e l o w . . S e n t e n c e ( 2 5 a ) i s d e r i v e d b y s e l e c t i n g a n d i n s e r t i n g k u t i

Z+ i n d i c a t i v e ? , a n d s e n t e n c e ( 2 5 b ) b y i n s e r t i n g k u t i / + s u b j u n c t i v e ? ,

i n t o T h i s y i e l d s t h e d e r i v e d u n d e r l y i n g s t r u c t u r e s i n ( 2 6 b ) a n d

( 2 6 c ) r e s p e c t i v e l y . I t i s e n t a i l e d i n ( 2 5 a ) t h a t M i d z i a c t u a l l y t o l d

l i e s b e f o r e t h e c o u r t ,

8 . op., c i t . p , 1 4 6

Midzi

B

v NP

nyara

N P V P

N P N P

taura nhema padare

(26)b

Midzi

nyara

k u t i 1+Indlc jJ"

V P

N P

taura nhema padare

( 2 6 c

Midzi

nyara

k u t i N P

N PN P

taura nhema padare

1A3

while in (25b) he definitely refrained from telling lies before the court,

It seems evident then that the difference in meaning between these two sentences is attributed to the different complementizers* These sentences differ only in the complementizers inserted into their complement clau™ ses. Further examples of pairs of sentences which differ in meaning by virtue of the complementizers inserted into their embedded sentences are given below*

(2?)a*!* Ndinotya^kuti nyaya^ iyi inobuda pachena ndi^ati^ndawana mari yokuripisa.(I fear that this affair will come into the open before I have the money with which to pay the fine*)

ii. Ndinotya kuti nyaya^ iyi ibude pachena ndisati* ndawana^

(1 am frightened in case this affair comes into the open before I have the money with which to pay the fine.)

b.i. Makanganwa kuti tsamba dzaenda mangwanani.(You have forgotten that the letters were taken in the morning.)

/ ✓ / / /iie Makanganwa kuti tsamba dziende mangwanani.(You forgot to post the letters in the morning*)

/ / / / / /e*i* Txri kumufungira kuti anouya mangwana.(We are expecting him to come tomorrow.)

.. / y , .. // / /1 1 . Txrx kumufungira kutx auye raangwana*(We are thinking of bringing him tomorrow*)

/ / / ✓ / / / s s /d0io Tendai anosema kuti vana vanodyira pane nhunzi.(Tendai is disgusted by the fact that the children have their meal at a place where there are flies.)

xi. Tendai anosema kutx vana vadyxre pane nhunzi.(Tendai is disgusted that the children should have their

meal at a place where there are flies.)

1/]./{.

e*i* Mashiri anoshora kuti Garwe akatiza aehisiya vana vake^* (Mashiri condemns the fact that Garwe ran away leaving his children behind*)

ii* mashiri anoshora Garwe kuti atize aehisiya vana vak/* (Mashiri condemns the idea that Garwe should run away leaving his children behind*)

f*i* Kuti Gani^anouya mangwana zvinondiWdza*

(That Gani is coming tomorrow pleases me*). . v, „ J i t t t 4 , 4 ^ / ,n e Kuti S a m auye mangwana zvmondifadza*

(For Sani to come tomorrow would please me*)It was mentioned above that complementizers could be regarded as consti­

tuting part of the rule features on verbs* But this line of approach

poses a problem in the case of some verbs* To this end consider the following data* Notice that in (28a) the complementizer in the compleme­nt clause following the verb -konzera is kuti /+ subjunctive/®

(2 8)a0io Kuti murume nemukadzi^ vagare^ vachirwa zvinokonzera kuti

varlkbane.(For a man and his wife to be always fighting will cause them to divorce each other*)

ii* Kana murume nemukadzi vakagara vachfrwa zvinokonzera kuti v/rJunban e *

(If a man and his wife are always fighting it will cause them to divorce each other*)

iii* Murume nernukadzi^ kugara vachirwa zvinokonzera kuti/ /varambane*(For a man and his wife to be always fighting will

cause them to divorce each other*) iv* *Kuti murume nemukadzi^ vagara vachirwa zvinokonzera

kuti varambane*

(*That a man and his wife are always fighting will

145

cause them to divorce each other*)tt ' / ' 7 f A * / * y, / / * y f 1 y A •v a Kana murume nemukadzi vagara vachirwa zvinokonzera kuti

Tar&ane,(If a man and his wife are always fighting it will cause them to divorce each other*)

The following sentences are the same as those in (28a) above save that

the object complement clause of the verb -konzera is introduced by kuti /A indicative/ in place of kuti £+ subjunctive/*

(28)b0i H *Kuti murume nemukadzi/ vagar/ vachirwa zvinokonzer^ kuti

varambana*ii0 *Kana murum/ nemukadzi^vakagara vachirwa' zvinokonzera

kuti varambana* iii* *Murume nemukadz/ kugara vachirwa zvinokonzera kuti

v / r W n a .iv, *ICuti murume nemukadzi vagara vachirwa zvinokonzera

kuti varambana*✓ / / / / / / / / / / /v* *Kana murume nemukadzi vagara vachirwa zvinokonzera

/ / /kuti varambana*

As above, the following sentences are the same as those in (28a) except/ / ft.that the object complement clause of the verb -konzera is introduced by

kana [*■ consecutive/ in place of kuti £+ subjunctive/*(2 8)c*i0 *Kuti murume nemukadzi/ vagare^ vachirwa' zvinokonzera kana

vakarambana 0iic *Kana murume nemukadzi/ vakagara vachirwa zvinokonzera

kana vakarambana*iii. *Murume nemukadzi^kugara vachirwa zvinokonzera kana

vakarambana *✓ / / / / / / ✓ / / / /iv„ *Kuti murume nemukadzi vagara vachirwa zvinokonzera kana

vakarambana*/ / / . / . y / / , ./ / ./ , / / ,v a *Kana murume nemukadzi vagara vachirwa zvinokonzera kana

vakarambana *

Ik6

In the following sentences the object complement clause of -konzera^ has

ku 1+ infinitive/ in place of kuti /+ subJunetive/.(28)dfti 0 Kuti murume nemukadzi^vagare vachirwa zvinokonzera

/kurambana*(For a man and his wife to be always fighting will causetheir marriage to break down*)

iia Kana murume nemukadzi^ vakagara vachirwa zvinokonzera /kurambana.

(If a man and his wife are always fighting it will cause their marriage to break down.)

xxi0 Murume nemukadzi kugara vachirwa zvinokonzera kurambana.

(same meaning as in (i) above)/ / / / / / / / / t t /iv0 *Kuti murume nemukadzi vagara vachirwa zvinokonzera

kurambana.Vo Kana murume nemukadzi^ vagara vachirwa zvinokonzera

/kurambana.(same meaning as in (ii) above)

The sentences which follow differ from those in (28a) in only one res­pect, namely, kana /+ participial/ replaces kuti /+ subjunctive/ in the

/ tobject complement clause of the verb -konzera./ / / / / / / / / / . / / /(28)e0i» *Kuti murume nemukadzi vagare vachirwa zvinokonzera kana

/ /varambana.ii. *ICana murume nemukadzi^ vakagara vachirwa zvinokonzera

kana varambana./ / / / / / / / / /iiio *Murume nemukadzi kugara vachirwa zvinokonzera kana

varambana^/ / / / ./ / / / / y / /,iv. *Kuti murume nemukadzi vagara vachirwa zvinokonzera kana

.a'rambka./ / / / ,/ / / i t J f t ,Vo *Kana murume nemukadzi vagara vachirwa zvinokonzera kana

Vciramban;

147

The rough underlying structure of the sentences in (2 8) is something like (2 9) "below*

(29)

konzera +kuti -t*subj +kana +consec„? Hku -i* in fin a-kana +partic*?

+pro

> J \ r \NP na NP V

gara vachirwa mukadzi

NP namurume

rambanamurume mukadzi

Nov/ let us see how, for instance, (28a 0i) is derived* No transformational rules apply when (or S^) Is cycled because at that stage it is not known which complementizer is to be inserted into it* In the second

cy&le, that is SQ , kuti /+ subjunctive/ is inserted into , and also into After this complementizer insertion and applying the other rules, the details of which do not concern us here, we eventually obtain sent­

ence (28a*i)* Nov; in order to derive the sentences in (28a*ii, iii, v) kana [ + consecutive/, ku /+ infinitive__/, and kana /+ participial/ respectively are inserted into S , but not into which continues to ^aYe kuti /+ subjunctive/ inserted into it*

As it has already been mentioned above the sets of sentences in (28b), (28c), (28d) and (28e) are the same as the set of sentences in (28a) except that kuti /+ indicative/, kana /-1- consecutive/, ku /+infin/ and kana /+ participial/ respectively have been inserted in place of

lmti £ subjunctive/ into S2 . The paradigms in (28) indicate (a) that

1k8f tthe verb -konzera may have in one and the same sentence both an absract

subject-NP and an abstract direct object-NP; (b) that the complement

clause of the subject-NP of this verb may have inserted into it eitherku /7 infinitive/;) or kana /+ participial/, or kana /+ consecutive/, orkuti /+ subjunctive/; and (c) that the complement clause of the object-NP

of this verb may have inserted into it either kuti /+ subjunctive/ or

ku /+ infinitive/ only. These observations are borne out by (a) the un-gramma t i c a 1 i t y of all the sentences in (28b ) 5 (28c) and (28e) in whichincorrect complementizers are inserted into and (b) the ungrammati-cality of the (iv) sentences in all these five sets (i*e, 28a - e)

because an incorrect complementizer is inserted intoIf it is insisted that complementizer insertion constitutes

part of the rule feature specification of a verb, then it has to be re-/ /fleeted in the complex symbol of verbs like -konzera that while

ku /+ infinitive/ or kuti /+ subjunctive/ may be inserted int$ a comple­ment clause which is in the subject NP position , as well as into a complement clause in the object NP position, kana /+ participial/ and kana /+ consecutive/ may be inserted only into a complement clause which is in the subject NP position of such verbs« I do not know how such rule features could be formulated in any current model. It is for this reason that kana /+ consecutive/ and kana /+ participial/ in the

tree diagram in (2$) appear with question marks, These two can only be inserted into S^, while kuti /+ subjunctive/ and ku /+ infinitive/ can be inserted into both and

We have a more or less similar situation in the sentences in (30)o In (30a) the complement clause of the direct object of the verb -ratidza has kuti /+ indicative/ as complementizer,

(30)a,io Kuti Deni^atakure tsaga iri zvinoratidza kuti akasimba,(For Deni to carry this bag shows that he Is strong,)

✓ / f / / *ir, Kana Deni akatakura tsaga iri zvinoratidza kuti akasimba,

(If Deni carries this bag it will show that he is strong,)

149

• • • ^ . / / / /1 1 1* Deni kutakura tsaga i n zvinoratidza kuti akasimba*(same meaning as in (i) above)

iv* Kuti Deni7 atakura tsaga iri zvinoratidza kuti akasimbaD(That Deni carried this bag shows that he is strong*)

V* Kana Deni* atakura tsaga iri zvinoratidza kuti akasimba,

(If Deni carried this bag it shows that he is strong*)In the following sentences the complement clause of the object of -ratidza have each kuti /+ subjunctive/*

(30)b*i* *Kuti Deni* atakure tsaga iri zvinoratidza kuti asimbe*ii* *Kana Deni7 akatakura tsaga iri zvinoratidza kuti asimbe*!

iii* "'-Deni kutakura tsaga iri zvinoratidza kuti asimbe*.

iv* *Kuti Deni*atakura tsaga iri zvinoratidza kuti asimb7*v* *Kana Deni*atakura tsaga iri zvinoratidza kuti asimb/*

In the sentences which follow kana consecutive/ is substituted for kuti /+ indicative/ in the complement clause of the object NP*

(30)c*i* *Kuti Deni7 atakure tsaga iri zvinoratidza kana akasimba*ii. *Kana Deni7 akatakura tsaga iri zvinoratidza kana akasimba*

iii. *Deni* kutakura tsaga iri zvinoratidza kana akasimba* iv* "Kuti Deni*atakura tsaga iri zvinoratidza kana akasimba*

Vo *Kana Deni*atakura tsaga iri zvinoratidza kana akasimba*Ku /+ infinitive/ substitutes for kuti /” indicative/ in the complement clause of the object NP in the following sentences*

(30)d*i* *Kuti Deni*atakure tsaga iri zvinoratidza kusimba*

ii* *Kana Deni7 akatakura tsaga iri zvinoratidza kusimba* iii* *Deni7 kutakura tsaga Iri zvinoratidza kusimba* iv* *Kuti Deni* atakura tsaga iri zvinoratidza kusimba*v* *Kana Deni* atakura tsaga iri zvinoratidza kusimba*

In the following sentences kana /+ participial/ replaces kuti / + indie,*/ in the complement clause of the object noun phrase*

(30)e*i* *Luti Deni atakure tsaga iri zvinoratidza kana akasimba*

150

iio *Kana Deni^ akatakura tsaga iri zvinoratidza kana akasimba,

iii0 -Deni^ kutakura tsaga iri zvinoratidza kana akasiraba^ iv, *Kuti Deni^ a^takura tsa^ga iri zvinoratidza kana To *Kana Deni^atakura tsaga iri zvinoratidza kana akasimb^u

The rough underlying structure of all these sentences is shown

in (3D

+N +CL8 +pro

Deni

ratidza +ku +infin,? j+kana +p^tic «,?/ +kana +coJ§ec„? ) +kuti +indie. I tf-kuti +sub

takura tsaga iriDeni

simba

If kuti /+ indicative/ is selected and inserted into both and we *”•* *”eventually come up with sentence (30a.iv), But as can be seen in (30a*i - iii and v) kuti /+ subjunctive/, kana /+ consecutive/, ku [ + infinitive/ and kana // participial/ respectively may also be inserted into , while only kuti /-i- indicative/7 may be inserted into The ungrammati- cality of the sentences in (30b - e) bears this out. So, as was stated above, in specifying the rule features of the verb -ratidza it has to be stated somehow that while kuti A indicative/ may be selected for the complement S in both the subject NP and the object NP, kuti /+ sub junc/7,

kana /-f consecutive/ 3 ku infinitive/ or kana / + participial/ may be

inserted only into a complement S of the subject NP with this verb.

The discussion in this section has brought into the open the

151

following problems with respect to the transformational hypothesis: _(a) complementizers are after all not devoid of semantic content as they have often been claimed to be; (b) the structural description of any

complementizer insertion transformation cannot not be limited to a complement clause, but must include the verb which that clause comple­ments; (c) a verb may have as its selectional restriction both an abst­ract subject NT with a complement S and an abstract object NP also vrlth a complement S* In some cases these complement sentences may have diffe­

rent complementizers as in (30aoi); (d) it is not uncommon to find that when two complementizers occur in one and the same sentence one comple­mentizer will be limited to one complement clause while the other will not be so limited® This is the case in (28a®ii - iii) for instance* The question which cries out for an answer is how are these facts to be in­corporated into rule features? All these shortcomings are a natural consequence oiff the transformational hypothesis* It is clear therefore

that the distribution of complementizers in Shona cannot be adequately described under the transformational hypothesis of complementizer insertion®

The Phrase Structure Hypothesis From the discussion in if*2 it is evident that complementizer

selection is the kind of phenomenon which is characteristic of subcate­gorization and that in fact some aspects of complementizer choice cannot be described within the framework of the transformational hypothesis without extending it in undesirable ways*

In an attempt to overcome the problems which arise from adopt­ing the transformational hypothesis as discussed above It is proposedhere to adopt Bresnan’s approach and regard complementizers as specified

gin deep structure® Bresnan called it the "phrase structure hypothesis"*

9* op* cite p* 300

152

She claims that there exists a phrase structure rule in every language having complementizers which permits complement types to be distinguish­

ed in deep structure* She set up the following rule for English:

(32) S ^ COMP + SThis rule seems to work equally well for Shonas According to this appro­ach the individual sentences in each of the pairs in (?) - (12) would have each its own underlying structures, For example, (7a) and (7b) would

have roughly the following underlying structures in (33) and (3k)

respectively,,

(33)

Masanjara V

Igona S.

A DVVPA sub3*/

Masanjara V1+ time7

S

enda mvura isati yanaya

Ok)

MasanJara V

[gona S

ADVVPNPI * consec *_/

Masanjara V

enda

/+ time/I "S

mvura isati yanaya

153

With this approach rule features become an unnecessary artifice v/ith

regard to complementizers,, The point to note here is that since the phrase-structure hypothesis specifies that complementizers exist in deep structure, verbs may be subcategorized for the type of complement­izer which they permit in their complement clauses0 For instance, in the case of the verb -gona in (7a) above we can say that it takes a comple­ment clause which is introduced by kuti /+ subjunctive/, and in the case of (?b) the same verb takes a complement clause which is introduced by

kana /+ consecutive/*,When I discussed the data presented in (28) and (30) certain

facts emerged which the transformational approach could not adequately handle0 For instance, there was the fact that in (28) the complement clause of the subject NP may have either ku /+ infinitive^ or kana

participial/, or kana /+ consecutive/, or kuti /+ subjunctive/, while the complement clause of the object NP is restricted to either ku /+infinZ- or kuti /+ subjunctive/ only* If we do not set up a node like COMP In deep structure, we would not be able to describe these facts by the use of the usual rule features or by any other known transformational device. The point is, if a complementizer insertion transformational rule oper­ated in (28a - e), it would have to be sensitive, not only to the rule feature on the verb, but also to the subcategorization of the verb -konzera. That is, ku /+ infinitive/ and kuti /+ subjunctive/ may be inserted into a complement clause which Is in either the subject or object position, while kana /+ participial/ and kana /+ consecutive/ may only be inserted into a complement clause which Is in a subject position. If complementizers are going to subcategorize verbs, the question is how is this to be formalized?^ moinen^ seems to be an open

10o One way, I suppose, might be to adopt the approach which is presen­

ted in rough below. Taking (7) and (13) as illustrative examples, the

15^question*

Let us examine more closely the sentences in (28a.i and iv) and (30a*i and iv) which are repeated here as (33) and (36) for conve­nience*

/ v / / / / / / / / / / / / /(33)a* Kuti murume nemukadzi vagare vachirwa zvinokonzera kuti/ /varambane.

t / / / / / / / / / / / /h*. *Kuti murume nemukadzi vagara vachirwa zvimokomzera kuti/ /varambane.

. _ V / / / / / / / /(36)a. Kuti Deni atakure tsaga iri zvinoratidza kuti akasimba* b* Kuti Deni" atakura tsaga iri zvinoratidza kuti akasimba.

These sentences present a problem. Notice that (33b) is ungrammatical*.In this sentence the complement clause of the subject NP has kuti /+ indicative/ while the complement clause of the object NP has kuti /+ subjiunctive/. But (36a) in which these two complementizers are rever­sed is perfectly well-formed, in (35a) the complement clauses of both the subject and

verb -gona in (7a -b) and (13b) will be regarded as one verb and thatin (13a) as another*. This sounds perfectly sensible since the two differ

in their meanings. The former may be entered in the lexicon as:

-gona Z+V5 -i- S ; /+ kuti + subjunctive/; / + kana -i-consec//and the latter as:

-gona /+V, + NP; /+_____ ku + infinitive/ S/,In the case of the first -gona this is to be interpreted as follows: this verb is followed by a complement. S.. This complement S must have either kuti /+ subj-unctive/ or kana /+ consecutive/, but not both..There should be no problem here because in the grammar it will be. speci­fied that complementizers are mutually exclusive as has already been stated* As for the second -gona, it occurs in the environment before

NP* Concerning what follows NP, namely, L v infinitive/ S / , this is

to be interpreted as follows: the complement of this verb is an abstract

NP which contains a complement clause*

155

the object have the same complementizer, namely, kuti /+ subjunctive/*

The same applies to (56b); both complements have kuti /+ indicative/.

The problem centres on (35b) and (36a). It would appear that there are interdependencies between complements which have an effect on the sele­ction of a complementizer. With regard to these two sentences (i.e. 35b

and 36a) it seems to be the case that when kuti /+ indicative/ occurs in the subject, kuti /+ subjunctiveZ cannot occur in the object. However, the reverse is not true as is indicated by the grammatical!ty of (36a)

in which the subject complement clause has kuti /+ subjunctive/ while

the object complement clause has kuti /+ indicatives/. Belov/ are further examples of sentences which illustrate this point. The sentences in (37) and (3 8) correspond to those in (35) and (36) respectively.

(37)a «i^ Kuti muswer/ makagara zvinoita kuti basa rireg/ kuenderera /mberio

(If you spend the day loafing there will be no progress in the work.)

ii* *Kuti maswera makagara zviuioita kuti basa rireg/ kuenderera /mberi.t t / / // / / / / / / f / fb.i* Kuti vaende kuniba iye zvino zvinovabatsira kuti vasaniwe.

(If they go home now it will save them from the rain.)/ / / // ✓ / , / , / / / . / / * /li. *Kuti vanoenda kumba rye z v m o zvinovabatsira kuti vasaniwe.

✓ f f / / / / / / / /c.i. Kuti uchovhe makasa zvinokurunzira kuti uite ubhinya.(If you gamble with playing cards it will lead you into being a savage.)

ii. *Kuti unochovha makasa zvinokurunzira kuti uite ubhinya., . / / / / / / / / / / / yd.i. Kuti unyarare zvinokunda kuti utukane naye,

(For you to keep quiet will be much better than quarrell­ing with him.)

ii. *Kuti unonyarara zvinokunda kuti utukane nay/** o \ / / / / / / / / / I I i f f !(3o)a.i. Kuti auye kuno zvinoreva kuti anoda mukadzi wake.

1 %

(For him to come here means that he loves his wife*)/ / / / / / / / / / / ( / / /ii, Kuti akauya kuno z v m o r e v a kuti anoda mukadzi wake*

(That he came here means that he loves his wife*)/ s / / / / / / / / / /hoi* Kuti vasiye imba yedu isina murthu zvinotaura kuti havatende*

(For them to leave our house unattended shows that they

are ungrateful*)/ / / / / / / / / /li* Kuti vakasiya imba yedu isina munhu zvinotaura kuti

/ /havatende *

(The fact that they left our house unattended shows that

they are ungrateful*)

If a node like COMP is not set up, then verbs cannot be subcategorized

for complementizers* As has already been pointed out this would necessi­

tate an extension of the transformational theory in undesirable ways*

For instance, a complementizer insertion transformation would have to

refer first to the verb in the higher S, But in some cases it would have

to refer also to a complementizer already placed on another complement

of the verb during the same cycle as was the case in (29) and (31) above,

Such interdependencies are however typical of subcategorization.

To this end consider a few sentences.in which the verb -ita (make) occurs,

(39)u0 Kuti mugare chete zvinoita kuti muve nenungo,,

(that - you sit - only -- it makes - that - you have - lazi­

ness = For you to be idle will make you lazy,)/ / t i t / / /b, Semeni akaita vana vake,

(Semeni - begot - children - her =

Semeni had her children,)e f .//,/.( . . . / * * I J Ac, Semeni akaita kuti vana vake vafunde,

(Semeni - made - that ~ children - her - they learn=

Semeni educated her children,)/ / / / / i f / / /d, *Kuti Semeni afunde zvakaita vana vake,

(•*That Semeni should learn made her children,)

157

e* Kuti Semeni afunde zvakaita kuti taka wake atengese mombe*

(that - Semeni - she learn - it made - that - father - her-

he sell - cattle - For Semeni to he educated her fatherhad to sell his cattle,)

We not®.that in (a) and (e) both the subject and the object are noun

phrases, of the type (N + S)} in (b) both the subject and the object are

human noun phrases, in (c) the subject NP is human while the object is an abstract NP of the type (N + S), and in (d), which is ungrammatical, the subject is an abstract NP of the type (N + S) while the object NP is human* What this signifies is that with certain verbs object noun

phrases other than those of the type (N + S) are not permitted when they

(i*e* the verbs) have subject noun phrases with an embedded complement clause* This is illustrated in (39<I) and ( 3 9 g )„

Those verbs which have multiple sentential complements offer

compelling evidence for setting up a COMP node in deep structures for they appear to be subcategorized for types of complements as was demon­strated above* There is yet another strong argument for setting up COMP

as a node in deep structure* This argument relates to the interaction of complementizers with other transformations*

One transformation which clearly interacts with complementizers is Conjunction Reduction* As an illustration of how this rule operates

consider (AO) and (Al)°- , / J / / / / / /(AO) Badza naMatenzeni vaenda Harare*

(Badza and Matenzeni went to Harare.*), N / // / / l i f t f(AD Badza a enda Harare 11a Matenzeni aenda Harare*

(Badza went to Harare and Matenzeni went to Harare*)

Through Conjunction Reduction (AO) is derived from (Al)* This rule, whichis cyclic, must precede on each cycle the cyclic transformation of gendercopying* The sentence^in (AO) and (Al) clearly illustrate this point*

Note that (Al) Is a paraphrase of (AO)* In (Al) Badza and Matenzeni each

158

controls a singular concord, namely, the underlined a® But in (AO) , be­cause the two noun phrases have now been joined together, they control as one entity the plural concord va which is also underlined®

The crucial argument here is that if complementizers were not distinguished in deep structure, but rather treated as being inserted transformationally, some grammatical conjoined sentences would not be generated® But if complementizers are specified in deep structure, then these sentences are generated with ease using familiar rules®

Assuming that complementizers are introduced transformationally, the way complementizer insertion and conjunction reduction interact depends largely on the tree configuration to which they are applied® One possible such structure is the one in (A2)*

(A2)

VP. VP,

With the transformational hypothesis the earliest point at which comple­mentizer insertion can apply in the configuration above is (or S^)

since it must be sensitive to the rule feature on the V in VP^ (or VP^) ® Row notice that conjunction reduction cannot apply before the SQ cycle®Let us illustrate this with an actual sentence®

/ / / / / / / / s / /(A3) Kana Beni akatuka baba uye kana akarova sekuru zvinonyadza*(If Ben scolded his father and if he beat his uncle it was a disgrace,)

The underlying structure of this sentence is something like (AAa)*

159

uye

NP VPVPNP.

_nyadzai/+kana*part±c/

VPNPNP VP

Beni NPNPBeni

tuka baba sekururova

( W b .

uye

VPNPVP

nyadzanyadza

NP VP

Jfcana _NP ^Z+partic/ |

Beni

VPBeni V NP

NPtuka baba

sekururova

160

( ) c

+pro

Jkana __ /+part±c/ FP

nyadza

Beni V

tuka

Jtana _ par tic/ F P VP

l r \Beni V FP

rova sekuru

(4^)d<

In P. uye NP.nyadza

+F +CL8 -5- proVP VP

VPBeni Beni NP

i ituka baba rova sekuru

No transformational rules apply when S^ and are cycled* As we move

on to (and also S 0) kana £+ participial/is inserted into (and also

161

^2 -cana /+ participial/ is inserted into (and also 3^) e This is shownin (AAb) 0 Now that it is known that S.. (and also S. ) will he in the3 Aparticipial mood, gender copying can apply to 3^ and also Noticethat the VP constituent in is identical to that in 3 ^ a At this stage

conjunction reduction applies, yielding the configuration in (AAc ), Thesubject NP of Sj is then deleted under identity with the subject NP of3^ o The head noun of NP^ and that of NP^ then together control the pluralconcord zvi as gender copying apolies to Sn 0 Note that apart from NP»“ ** -L 2the rest of the node disappears as a direct consequence of the con­junction reduction rule, Since, after conjunction reduction has applied,SQ no longer branches, it is deleted by the- convention which delete any

11S node which does not branch. This results in (AAd) • After applying

later rules which are not crucial to the present discussion we eventually

obtain sentence (A3)®>In the case which has Just been considered above complementizer

insertion preceded conjunction reduction. However, t h e r e are instances in Shona when conjunction reduction precedes complementizer insertion in a derivation, as for example in the structure like that sketched in

(A3) below,

U5)

NP VP

V NP

S.

s

11* For a discussion on S-pruning see John R„Ross in "A Proposed Rule

of Tree-pruning” in Modern Studies in Pnglish, David A,Reibel and

Sanford A*Schane (editors).

162

In this diagram the structural description of conjunction reduction is

met at S , hut the structural description of complementizer insertionreduction

is not met until SQ is cycledo Here then is a situation where conduction

will obviously have to apply before complementizer insertion in a derivation,,

As an illustration consider the derivation of (46)«,(4 6) Jeni^ anoda kuziva kana sadza uye kana maburo zvabikwa*nevaroora0

(Jeni wants to know whether the sadza and whether the sweet

beer have been, prepared by the daughters-in-law0)The structure which underlies this sentence would look something like (47a) below,

(47) a *

J e m V

— da _ ^ku+infin/

[+N I HP-i-CL 8 •{-pro Jeni V

I_ ziva _ N

/kana+partic/ j~+N +CL8

1

PreS H P VP

H P passive

+prq

PreS NP VP

passivevaroora

V NPvaroora]

bika sadza bika maburo

163

(47)b

J eni V

da

VPNP+N+GL8+pro Jeni

uye

VPVP

I / Asadza V

Ibikwa

I / Af V

I

PP maburo V PP

nevaroorabikwa

nevaroora

(J+7)c* S

J eni V

J e m V

_ da _ ^ku+infin/

zxva

VP

sadza

uye NP

maburo bikwa nevaroora

/ 11 ppsadza maburo |bikwa

nevaroors

When and S. are cycled the passive rule is applied yielding (47h)»We notice that complementizer insertion cannot a-only on iS7 S. because~ 3 or ^of the nature of the next higher node* namely, However in the second

cycle? i*e* the structural description of conjunction reduction ismet* Applying this rule the result is the intermediate structure in (4?c). The node then disappears by the pruning convention mentioned above„

At this point complementizer insertion can apply and the tree which re­sults after its application is (47d)° After applying later rules the sentence which ultimately results is (48)0

(4 8) Jeni anoda kuziva kana sadza uye maburo zvabikwa nevaroora*(Jeni wants to know whether the sadza and the sweet beer have

been prepared by the daughters-in-law*)

Notice that this sentence, though grammatical, is not the same as the one in (46) above* There is some difference in connotative meaning bet­ween (4 6) and (4 8) * In (4 6) the sadza and the maburo seem to be more

165focussed than they are in (l8)0 The point however is that there is no way in which complementizer insertion can produce (16) from the structure

in (15) and (17)® If we cling to the transformational hypothesis we could, I suppose9 allow complementizer insertion to apply when is cycled as in (17d) and set up a later rule which optionally distributes a complementizer preceding a conjoined structure over the conjoined noun phrases® Such a rule would allow us to derive (19) from (17d)0

(19)

Jeni

_ da /ku-fin fin®/

-{-pro

zxva NJ -

•{•pro

NP uye NP bikwa/ A nevaroora

Jkana _ N kana ___ N1+ parti c®/ j / j - par tie®/

sadza maburo

This distribution of the complementizer over the co-ordinate structure in (I7d) would eventually result in the derivation of (16)® Allowing such a rule creates problems elsewhere® For one thing it cannot be said to be optional because, as has already been pointed out, (16) and (18)

have different pragmatic interpretations,, If we allow it to be an obli­gatory rule such a solution would permit (50) to be transformed into

166

(5 1) which is an ungrammatical sentence*

(50) Ndinoziva kuti Deni^naFani^vakasangana.(I know that Deni and Fani met.)

(51) *Ndinoziva kuti Deni^ nekuti Fani^vakasangana.(*I know that Deni and that Fani met.)

In (50) we have a reciprocal predicate which cannot be derived through

conjunction reduction* Thus (50) cannot be said to derive from (52)./ v i t / A > / / / / / / / / /(52) *Ndinoziva kuti. Deni akasangana nekuti Fani akasanganae(*I know that Deni met and that Fani met.)

The conclusion to be drawn from this is that this ad hoc complementizer distribution rule would have to apply only in those cases where conjun­

ction reduction will have applied,, But the drawback is that after con­junction reduction has applied the conjoined noun phrases can hardly be

IPdistinguished, if at all, from the non-reduced cases of NP conjunction* The crucial point here is that if the complementizers in (4 6)

are present before conjunction reduction applies, their derivation is accounted for without any difficulty. But with the transformational hypothesis there does not seem to be a satisfactory way in which (46) could be derived. The argument concerning the interaction of compleme­ntizers with conjunction reduction appears to prove beyond doubt that complementizers ought bo be specified in deep structure rather than introduced transformationally.

Questions in Shona may be embedded in higher sentences just as in many other languages. The following are examples of some common types of simple unembedded questions in Shona. The interrogative formatives are underlined in each question. (For embedded and non-embedded questions

12. For a discussion on conjuncts see G.Lakoff and S„Peters in "Phrasal Conjunction and Symmetric Predicates" in Modern Studies in English,David AoReibel and Sanford A.Schane (editors), p.113 - 142.

167

see also ‘ 2*5*1 and 2*5*2), . / / / / /(53)&* Beni akadzoka here?

(Beni - he came back - question = Did Ben come back?)/ / /b* Changufu anobva kupi ?

(Changufu - comes - from where = Where does Changufu come

from?)✓ / / / / / /c* Hazvina akaenda Harare neyi?

(Hazvina - went - Harare - with what =

With what did Hazvina go to Harare?)f t / td* Nzou yakaita seyi?

(an elephant - be - like what =

What does an elephant look like?)/ / *56 /e* Beyi wakatiza?

(v:hy - you ran away - Why did you run away?) r. •f« Swmza m o d y a chiyi?

(a giraffe ~ eats - what = What does a giraffe eat?)c ' / *og. Sara arova ana?(Sara beat whom =: Whom did Sara beat?)

When these sentences occur as indirect questions, that is when they occur as embedded questions, the complementizer employed is usually kuti

/+■ indicative/ as the sentences in (5k) show*(5k)a* Ndinoda kuziva kuti Beni akadzoka here.

(I want to know whether Ben came back.) b* Ndinoda kuziva kuti Changufu anobva kupi*

(I want to know where Changufu comes from*) c* Ndinoda kuziva kuti Hazvina akaenda Harare neyi*

(I want to know what means Hazvina used to go to Harare*)/ / / / / / / /d* Ndinoda kuziva kuti nzou yakaita seyi*

(I want to know what an elephant looks like*)/ / / / / / / /e* Ndinoda kuziva kuti seyi wakatiza*

(I want to know why you ran away.)

168* * f f . . y , / ,. yf, Ndinoda kuziva kuti swiza m o d y a chm.

(I want to know what a giraffe eats,.)t i l l I / Ig. Ndinoda kuziva kuti Sara arova. a m ,

(I want to know who Sara beat,)

Although there are many verbs which permit a complement clause which

has kuti /+ indicative/* * not every one of them will allow an embedded question,.

/ t / t / / / / / / ✓ , /(55)sui* *Tichaona anodaira kuti zvinhu zvese zvichanaka here,(^Tiehaona believes whether everything will be all right.)

ii. *Baba anofungidzira kuti mwara^ wake anosvika nhasi here.(*The father expects, whether his child will arrive today.)

I f f / /iii, *Mwana, akanganisa kuti atuka arnai here,

(*The child made a mistake whether she scolded her mother.)I f f / / / / / / / / / /iv. *Taka'. ari kushora kuti VaMago vakatiza mhuri yavo here.

(*Taka is blaming whether Mr.Mago deserted his family.)/In the sentences in (55a) the interrogative formative: here is question­

ing the embedded clause. But if this interrogative formative is madeto question the whole sentence, these sentences become fully well-formed.

I f t / / / / / f / / f(55)b,i, Tichaona anodaira kuti zvinhu zvese zvichanaka here?(Does Tichaona, believe that everything will be all right?)

/ / / / / ✓ / /ii. Baba anofungidzira kuti mwana wake anosvika nhasi here?(Does the father expect, his child to arrive today?)

if t i f f Iiii. Mwana akanganisa kuti atuka amai vake here?(Did the child make a mistake by scolding her mother?)

I // i i t / / / / / / fiv. Taka ari kushora kuti VaMago vakatiza mhuri yavo here?(Is Taka deploring the fact that Mr.Mago deserted his family?)

That verbs can be subcategorized according to whether they can permit

indirect questions or not is a well known phenomenon.As was pointed out earlier one thing peculiar to complement­

izers is that they are mutually exclusive. The following sentences

contain

169

two complementizers each, and these are underlined*

(56)a, *Handizive kuti kana akauya.(*I do not know that whether he came*)/ / / /Vanoda kuti kuenda kumusha.

(*They want that to go home*)

c. *Jojo akati^ kana j2f aizotora ush/ nesimbal(*George said whether that he would seize the chieftainship

by force,)But the sentence in (57) below appears to be a counter-example to this

claim*(57) Handizj^ve kana kuti achasvxka pano her e nhasi*

(I do not know whether (that) he will arrive here today =I do not know whether he will arrive here today,)

In this sentence we have kana and kuti immediately following each other.As was stated earlier kana as a complementizer is followed by a clausein either the participial mood or the consecutive mood, while kuti isfollowed by one in either the indicative mood or the subjunctive mood.In (57) the complement clause Is in the indicative mood, clearly indica­ting that the complementizer here is kuti /+ indicative/.. That being the case what then is the status of kana in this sentence? Note that kana here can be left out without altering the meaning of this sentence,

(58) Handizive kuti achasvxka pano here nhasi*(I do not know whether he will arrive here today,)

Here are further examples of sentences of this type,. The (i) sentences

below correspond to (57) and the (ii) sentences to (58)«/ / / / / / / / / / .(59)a.*i* Paakasangana neshumba akashaya kana kuti oita seyi*

ii, Paakasangana neshumba akashaya kuti oita s<fyi,(When he came face to face with a lion he was at a loss

as to know what to do.)/ / / / / / / /b„i, Hameno kana kuti Kondo akadyei; aonda chaizvo*

170

ii. Hameno kuti Kondo akadyei; aonda chaizvo.(I do not know what Kondo ate; he has lost a great deal of weight,)

/ / / / / /c.i. Zvandakanzwa my ay a iyi ndakabva ndashaya kana kuti ndoitei./ / / / / / / / / /ii. Zvandakanzwa my ay a iyi ndakabva mdashaya kuti ndoitei.

(When I heard this story I did not. know what to do.)In the following sentence (60) the second kana does not seem, to be. a complementizer but is rather a conjunction of some kind.

(60) Handizive^^ kana kuti achasvika pano her a nhasi kana kutit / *haachasvika.

(I do not know - whether - that - he will arrive - here -

today - or - that - he will not arrive =I do not know whether or not he will arrive here today.)

While the first kana in this sentence is optional as we saw above, the second one is obligatory.

(61)a. Handizive kuti a^chasvika pano here nhasi kana kuti

haachasvika.f t , v / I f I / / /b. *Handizive (kana) kuti achasvika pano here nhasi kuti

/ I /haachasvika.It seems quite evident that the second kana is serving to co-ordinate the two embedded complement clauses. The status or function of the first kana however still remains unclear,. If the argument that comple­mentizers in Shona normally have moods associated with them is valid, this kana then is not a complementizer formative since none of the moods associated with it is reflected in the complement clause. More­over, the fact that its occurrence in such sentences is optional strong-

15ly suggests that it is not a complementizer, as normally complementi­

zer formative© are

13. I say ''normally11 because© there are a few cases when a complement­

izer seems to be deleted from an embedded sentence (see 6.1.1.4)*

171

either obligatorily present in the case of embedded sentences or obliga-

torily deleted in the case of non-embedded sentences,, Notice also that

kana as a lexical item has a multiplicity of meanings; among them are:when; if; although;even if; or; whether ----- or. In the absence of anycompelling evidence that it is- a complementizer formative in the case

under discussion, I take the view that it is not a complementizer here although the question is probably an open one. In this case the claim made on p. 68. , namely, that complementizers are mutually exclusive,

holds,.We observe also that where complementizers consist of an overt

formative and a mood, with the exception of ku and zva, the formative may occur in either a clause-initial position or a position immediately after the subject NP* It seems that the occurrence of the subject NP of the complement clause before the complementizer formative can be

accounted for by an optional transformational rule which interchanges

them. The examples in (62) and (63) illustrate the occurrence of comple­mentizer formatives in clause-initial position while those in (64) and (63) illustrate their occurrence in a non-clause-initial position. The complementizer formatives are underlined in each case.

(62) subject complementation;a 0 Kana madzimai akatukana pamberi pamambo zvakanyadza.

(If the women scolded each other in front of the chief itwas a disgrace.)

/ / / / / / / / /Kana sekuru akasvika kunze kwaenda hazvichabatsira nokuti/ / / / / / / i fbasa racho rinofanira kuitwa kuchiri kuchena.

(If uncle arrives late in the night it will not help us because the ritual has to be done before it is dark.)

14a This applies if one goes on the assumption that every sentence has

a complementizer in its deep structure, (see also 4*4 below.)

172

/ f i t J / I tCo Kuti nyika yakatenderera zvmozivikanwa•

(That the world Is round is known*)fi i / / /do Kuti mwana achate zvinofadza vabereki*

(For a son/daughter to wed pleases his/her parents,)

(63) object complementation;/ i f / / / / . /a» Ndabvunza kana Farai akadzoka kuWedza0

(I asked whether Farai came back from Wedza,)✓ f i f f f / fb 0 N ’anga iye ichazvibata chete kana Gondwe akasaidzira mudzimu

/ /wacho kwatirio(That witch-doctor will be able to detect it even if Gondweincites the spirit to turn against us,)

/ / t f / i /Co Tinoziva kuti nyika yakatenderera*

(We know that the world is round*)f i t i f f ( f t /do Hubereki wose anoda kuti mwana wake achate,

(Every parent wants his/her son/daughter to wed,)

The meanings of the sentences in (6k) and (63) below are the same asthose of the corresponding sentences in (6 2) and (63) above*

(6k) subject complementation with complementizer formative in anon-clause-initial position;

f i . i t J i t , /a* Madzimai kana akatukana pamberi pamambo zvakanyadza*/ t t f f t s s fb 0 Sekuru kana akasvika kunze kwaenda hazvichabatsira nokuti

basa racho nnofanxra kuitwa kuchin kuchena*/ i f f t / / /c, Nyika kuti yakatenderera zvinozivikanwa*

d, Mwana kuti ^fchat/ zvino fadza vabereki0

(63) object complementation with complementizer formative in a non­clause-initial position;

f / f i t i * f 15a* Ndabvunza Farai kana akadzoka Wedza*

15® This sentence is ambiguous* Farai can be the direct object of thef t t fverb -bvunza or it can be the subject of the predicate akadzoka Wedza,

Here we are concerned with the latter reading*

173/ / / / / / f tb* N'anga iye ichazvibata chete Gondwe kana akasaidzira mudzimu

wacho kwatiri*f i t f i t /c* Tinoziva nyika kuti yakatenderera*

/ / / / / / / / i td* Muberekx wose anoda mwana wake kuti achate*In the case of the complementizer formatives ku and zva9 if the subject NP of the complement clause is overtly expresseds they will always occur after it (i0ee subject NP)*

(66) Ndyire chaiye kuuya kuzotiona mnotishura*(For Ndyire to pay us a visit is an ill omen for us,)

(67) Mapondera zvaadzoka akaisa uso pasx zvxnoreva kuti sekuru varwara chaizvo.(Since Mapondera*s face was downcast when he returned it means that uncle is seriously ill,)

Note that complementizer formative ku preceded by a subject NP does not seem to occur in a complement clause which is in an object position? as is indicated by the ungrammaticality of (68),

(68) ^Ndinoda^ mwana/ kuenda*(I want the child to go*)

See also the (ii) examples in (170) in Chapter Two*Notice also that the complementizer formative kana in kana

/+ consecutive/ may be optionally deleted without effecting any semantic change.

(69)a„ Seki/ru akasvika kunze^ kwaenda hazvichabatsira nokuti basa/ f t * t / f f tracho rinofanira kuitwa kuchiri kuchena, (cp 62b and 64b)

/ / t t i f f tb* N'anga iye ichazvibata chete Gondwe akasaidzira mudzimu/ twacho kwatiri* (cp 63b and 63b)

In the case of the other complementizer formatives this optional deletion rule may not apply*

(70)a, *Madzimai akatukana pamberi^pamambo zvakanyadza* (cp 62a)/ / / / f f t !b 6 *Nyika yakatenderera zvinozivikanwa, (cp 62c)

m

c® *Mwana achate zvrnofadza vabereki® (cp 62d)/ / / / ✓ / f / . ^d® ^Ndabvunza Farai akadzoka kuWedza. (cp 63&)./ / A ^ / J f f n

f® *Mubereki wose anoda mwana wake achate® (cp 63d.)Although (70f) is unacceptable, at least in my idiolect, (71) below, which has a linear sequence of words similar to that of (?0f), appears to be acceptable, though marginally®

How is this discrepancy to be accounted for? The answer seems to lie in the; fact that while (63d.), from which (7 0 f) would derive, expresses a mere wish, (72) on the other hand, from which (71) derives, signifies much more than a wish; it is almost a command® There may be another

-n : reason which is not immediately apparent®

l\. * A Complementizers and Non-embedded sentencesIt is claimed in this study that every sentence, embedded or

non*-embedded, starts off as S in Shona® The implication is that every sentence in Shona has a complementizer which serves to introduce it thus

The important thing to observe is that in the case of the topmost S the complementizer formative will be deleted obligatorily by a convention® This convention may be tentatively stated as follows: delete the comple­mentizer formative from just in case is the topmost S in the deri­vation® It must be borne in mind that this convention will only be in­voked after mood incorporation has applied otherwise the predicate will

be left without a mood® That such an approach is necessary becomes ob­

vious if we bear in mind that even the topmost S in a complex sentence

' J 1 . * t I f(71) Ndinoda Kerr asvike pano mangwana (1 want Keri to arrive here tomorrow®)

COMP S

175

must be in one mood or another, Besides, if mood appears in deep struc­ture under COMP, as has already been stated above, it follows that there

must be COMP also to introduce mood into the topmost sentence. The alter­native is to say that mood is introduced into the topmost sentence, not by COMP, but by some other rule or process. If the latter alternative (irrespective of its formulation) is adopted, then we shall have a strange situation in which mood is introduced by COMP in some cases but

by something other than COMP in other cases,, It is clear that by adopting the second alternative our grammar will fail to capture a linguistic generalization, namely, that mood is being introduced into the predicate phrase in either case. Consequently the second alternative will result

in a weaker grammar. With the first alternative this linguistic genera­lization is captured. It may well be that the occurrence of COMP in the topmost sentence as well as the convention suggested above is universal,

to 5 SummaryIn this chapter I have tried to show that in Shona (a) comple-

mentizers have semantic content and t«her»e they should be specified in deep structure; (b) complementizers subcategorize verbs although it is not clear how' this is to be done; (c) these complementizers are mutually exclusive, that is, they are incompatible within the same minimal clause; and (d) complementizers are basically clause-initial, but nonetheless in most cases by a transformation the complementizer formative may per­

mute with the subject NP of the complement clause.

176

5.0 TRANSFORMATIONAL PULESIn this chapter I should like to discuss the following trans'

formational rules; which were mentioned in previous chapters:Rule (1) PassiveRule (2) Pseudo-passiveRule (5) Agent Deletion

Rule U ) Reflexive

Rule (5) Mood IncorporationRule (6) Complementizer Placement

Rule (7) Gender Copying

Rule (8) Object RaisingRule (9) Subject RaisingRule (10) Equi-NP Deletion

Rule (11) Extraposition

Rule (12) Pro DeletionRule (13) Object Anaphora

Rule U k ) Object Focussing

Rule (15) Object Copying

Rule (16) Noun. prefix (segment) rule

Rule (17) Subject PostposingA detailed account of these transformations would he such an enormous

undertaking that it would run to a volume by itself. Such a treatment

falls outside the purview of the present study; for the purpose in hand, a sketch of each of these transformations will suffice. This study has been carried out within the framework of a cyclic grammar. The concept of a cyclic grammar is discussed in subsection 5.18 in this chapter*

5*1 Passive (cyclic)Like many other languages Shona has among its various types

of sentence passive as well as active sentences* That these two types

of sentence are somehow related is well known. Almost every Shona grammar

177

that has been. produced to date describes the passive construction in one way or another* What is unclear however is precisely what the nature of the relation between passive sentences and their corresponding active sentences is* To this end consider the following sentences;

(1)a, Mapfeni akauraya mbada,(Baboons killed a leopard,)

b* Mbada yakaurayiwa nemapfeni*(A leopard was killed by baboons,)

(2)a, Kuti akwire” bhasikoro zvakanetsa Zuze,(that - he ride - a bicycle - troubled - Zuze =It was a bother for Zuze to ride a bicycle,)/ / / / / / ✓ / /b* Zuze akanetswa nekuti akwire bhasikoro,

(Zuze - was bothered - by that - he ride - a bicycle =It was a bother for Zuze to ride a : bicycle.)

As can be observed in these sentences, roughly what the passive trans­formation does is first to interchange the subject noun phrase with the object noun phrase, secondly to introduce the passive element /w/ into the verb, and thirdly to introduce the agent formative /NA/*

The passive transformation will operate on a structure of the type given below:

NP^ - AUX - V - NP ,The following conditions however must also be satisfied, First, NP^ and N?2 must not be referentially identical,. In (3a) below the object HP Tsitsi^ is formally and referentially identical with the subject NPTsitsi^, hence (3b) is ill-formed,^

, * ./ / i t /(3)a, *Tsitsx^ akarova Tsitsi^.(■^Tsits^ beat Tsitsi^.)

/ / / / /b, *Tsitsx^ akarohwa naTsitsi^.

1* For more discussion on referential identity see P.M,Postal in Cross-over Phenomena, Chapters 1 and 2.

178

(*Ts±tsi-^ was beaten by Tsitsi^.)Nevertheless it is possible for NP^ and NP2 to be identical in form while not being co-referential* In such a case the passive rule can apply readily as is illustrated in (ifb) below,

(Zf)a, Tsitsi^ akarova Tsitsi^,(Tsitsi^ beat Tsitsi^ i*e, the two are different persons)

/ / / / Ib, Tsitsi^ akarohwa naTs.itsi(Tsitsi^ was. beaten by Tsitsi^.)

Secondly, the verb phrase in the structural description of this rulemust be of the type that allows the passive to operate, There are certainverb phrases in Shona which, though they contain noun phrases as comple-

2ments, do not permit this rule to operate. Examples of such verb phra­ses can be seen in the following sentences:

/ / / /(5)&, Mwana uyu anenge rema*(This child seems to be a fool,)

/ / / / ,b, Tsaga iri rrnorema zana remapaunziL.

(This lag weighs lOOpounds.)If the sentences in (5) are put into the passive the results are the sentences in (6),

(6)a, *Rema rinengewe nemwana uyu,* / . yb, *Zana reraapaunziL rxnoremwa netsaga i n ,

Since there are very few such verbs in Shona, they can be marked in the lexicon as not permitting the passive rule to operate. Those which allow

3this rule to operate will then appear unmarked.There is yet another type of verb phrase which will not permit

2, cp behaviour of middle verbs in English,3, See G,Lakoff in Irregularity in Syntax for rule specification in

the lexicon.

179

the passive to apply eiSEen though it appears to consist of a verb and a noun phrase* Such is the verb phrase in which the object noun phrase constitutes an inalienable part of the nearest noun phrase on its leftas is shown by the sentences in (7) - (10).

. / / / /(7) Ngoro yabva vhiri.(a wagon - came off - a wheel The. wagon lost a wheel.)

i / / /(8) Momb.e. yakatyoka gumbo..(a cow - broke - a leg rr The cow had its leg broken.)

, x f / / 4 / / /(9) Muvhimi akapfura nzou dundundu.(a hunter - shot - an elephant - a chest =

The hunter shot an elephant in the chest.)t / / / f l(10) Fani akacheka Sara, nzara.

(Fani - cut - Sara - finger-nails =- Fani cut Sara’s finger-nails..)

The. verb, phrases in both (7) and (8) each consist of a verb and a nounphrase, while in both (9) and (1.0) they each consist of a verb and twoobject noun phrases. Vhiri in (7) and gumbo in (8) constitute inaliena­ble- parts of the subject noun phrases ngoro and mombe respectively. , ■In this case the passive rule will not apply.

' / < * /(11) *Vhiri rabviwa nengoro,/ / f(12) * Gumbo ratyokwa nemo mb e..

The sentences in (9) and (10) can be turned into the passive by inter­changing the first object noun phrase with the subject noun phrase. Thiswill result in the following sentences which are well-formed;

t / / / / / / /(13) Nzou yakapfurwa dundund^t nemuvhimi*(An elephant was shot in the chest by the hunter.)

l i t * / /(IZf) Gar a akachekwa nzara naFani.(Sara’s finger-nails were cut by Fani.)

On the other hand, because the second object noun phrase in (9) and (10)forms an inalienable part of the first object noun phrase, these two

sentences cannot be passivized by interchanging the second object noun

18o

phrase with the subject noun phrase as is shorn by the ungrammatical!ty

of the sentences below*(15) * Dundundu" rakapfurwa nzou n/muvhimiC

f / / f f /(16) *Nzara dzakachekwa Sara naFani.In (?) ^ (10) it is assumed that their rough deep structures are respect­

ively (a) and (b) below:(a) s (b) g

VPNP. NP. VP

NP NP

However this assumption seems to be counter-intuitive, especially when/

we see that in (7), for instance, it was only the vhiri which came offand not the whole ngor o as such. Accordingly the logical subject of the

/ /verb -bva here id vhiri and not ngoro* In (10) what was cut were the/finger-nails only and not the whole of Sara as such. The logical object

/ / / inoun phrase of the verb -cheka is in this sentence nzara and not Sara*If the logical subject noun phrases in (7) and (8) are made

the grammatical subjects also, and if the logical object noun phrases in (9) and (10) are also made the grammatical objects, the following sentences- which are grammatical will result:

(17) Vhrri rengoro rabva.(The wagon's wheel came off.)

(18) Gumbo remombe rakatyoka.(The cow's- leg got broken.)

t ' t / / / ,/ /(19) Muvhimi akapfura dundundu renzou*(The hunter shot an elephant's chest*)

/ / / / / ✓ /(20) Fani akacheka nzara dzaSara.(Fani cut Sara's finger-nails.)

Note that the subject noun phrases in (7) and (8) and the first object

181

noun phrases in (9) a-&d (10) have become possessives in the process* It

is possible to regard the phrase-markers of the sentences in (17) - (20)

as the structures which underlie the sentences in (7) ~ (10) respectively*

This; appears, to be a logical and natural thing to do* Nonetheless there

is a problem in adopting this approach. The sentences in (?) - (10) can

be derived from the corresponding structures which underlie those in

(17) - (20) only when the head nouns of - : the subject noun phrases in

(1 7 ) and (1 8 ) and the head nouns of the object noun phrases in (1 9 ) and

(20) exhibit an inalienable relationship with the accompanying possessor

in that subject or object noun phrase. If this relationship goes not

obtain, this transformation will block as illustrated in the sentences

which follow.* * t t(21.) a. Penzura yemwana yatyoka*

(The child's pencil got broken.)f t fb* *Mwana atyoka penzura.

(*The child got broken a pencil.)f . f / * y.(22)a. Madomasi. asekuru akatsvuka.

(Uncle's tomatoes have become red.)/ / / / /

b* *Sekuru akatsvuka madomasi.

(*Uncle became red tomatoes.)

In the noun phrases penzura yemwana and madomasi asekuru the head nounsf t /penzura and madomasi do not constitute an inalienable part of mwana and

sekuru respectively, hence the ungrammaticality of (21b) and (22b).

If this analysis is correct, as I believe it is, the feature

1+ inalienable/ must then be built in somehow in the subcategorization/ /statement of such verbs as -tyoka and - 1svuka. Since this presentation

follows the generative transformational model and the standard theory

in this model as currently formulated does not seem to permit this ex­

tension this presents a difficulty. Owing to the complexities involved

here I leave it as an open question as to which is the more viable of

182

the two suggested analyses. In this study I have adopted the former analysis although it appears to be counter-intuitive.

Now note that in each of the sentences in (23) and. (24) below there are also two object noun phrases just as in each of the sentences in. (9) and (10), These object noun phrases do not however exhibit the part-and-whole relationship that was seen, in respect of the object noun

phrases in (9) and (10),f f / ,, * -(23) Arabuya vapa muzukuru mabhanzx.

(The grandmother gave her grandchild some buns,)

(24) Rowesai^kakanda Chipo bhora,(Rowesai - threw - Chipo - a ball =Rowesai threw a ball to Chipo,)

Each of the sentences in (23) and (24) can be passivised by interchang­

ing either of the two object noun phrases with the subject noun phrase.The resultant strings will be perfectly acceptable Shona sentences,

/ / / /(23)a. Muzukuru apihwa mabhanzi naambuya.(The grandchild was given some buns by its grandmother,)

f t / /b, Mabhanzi apihwa muzukuru naambuya,(Some buns were given to the grandchild by its grandmother.)

, ^ / / / / / //(2b)a, Chipo akakandwa bhora naRowesai,(Chipo had a ball thrown to her by Rowesai.)

/ / / / / / /b, Bhora rakakandwa Chipo naRowesai.(A ball was thrown to Chipo by Rowesai.)

Notice that it is possible for the two noun phrases to be

distinct in form but to be fully or partially co-referential. In such

cases the passive will not apply. To this end consider the sentendes which follow:

/ / / / f(27)a. Posani akabata bhora.(Posani caught the ball.)

b. Bhora rakabatwa naPosanx.(The ball was caught by Posani.)

183

(28)a.*Shumba yakauraya mhondoro (*A lion killed a lion.)

/ / / / /b* *Mhondoro yakaurayiwa neshumba(*A lion was killed by a lion.)

/ t / / / / /

(Sara beats up her children.)r r f f f ( fb. *Vana vake vanorohwa naSara.

(*Her children (i.e. Sara's) are beaten up by Sara.)

The two noun phrases which are to interchange in (27a) are different both in form and in reference, hence the passive is permitted as (27b) shows. In (28a) the two noun phrases to interchange, though distinct in form, are nonetheless co-referential, hence (28b) is ill-formed. Note that whenever the subject and the object in a sentence are referentially identical the latter is obligatorily reflexivized. This accounts for the ungrammaticality of (28a). Now in (29a), even though the subject noun phrase and the object noun phrase are distinct from each other in form,there is however an element in the object noun phrase which refers to

/ / /the subject noun phrase Sara. This element is -ke in vake (her). Forthis reason the two noun phrases in (29a) cannot be said to" be strictlynon-co-referential, hence the passive will not operate as is shown in

/ / /(29b). If, howevej?, vake in (29a) does not refer back to Sara, the sen­tence will passivize since the two noun phrases in question will then be strictly non-co-referential.

This formulation of the passive requires comment* Its structural descrip-

Having discussed the structure which underlies the passive

construction we can now proceed to formalize it.Rule (1) Passive

X - passive - NPX - AUX - V - (NP£) - NP^ - Y

1 2 3 k 5 6 7 8

1 0 7 if 6 0 8 NA + 3

184tion has been formulated with two object noun phrases* Notice that NP^ is bracketed. This is meant to accommodate the derivation of (2?b) and (30b - c). In deriving (27b) and (30b) term 6 will be null., while in order to get (30c) term 6 will be the noun phrase nyoka.

(30)a, Vana vauraya nyoka kudanga.(The children killed a snake at the cattle-pen,)

i f f f t tb, Nyoka yaurayiwa kudanga revana.(A snake was killed at the cattle-pen by the children.)

c. Kudanga kwaurayiwa nyoka 11 e vana

(At the cattle-pen was killed a snake by the children.)In sentences like those in (31a) and (32a) in which an adverbial occurs between the verb and its object noun phrase the adverbial is considered to be outside the verb phrase at the time when the passive rule applies*

(3 1)a. VaMatenzeni^ vakabika kuchiri^kuchena sadza revana vemukuwasha* (Matenzeni prepared before It was dark the sadza for thechildren of her son-in-law.)

/ i f f / i f f / s / f /b. Sadza revana vemukuwasha rakabikwa kuchiri kuchena naVaMate-/ /nzeni.

(The sadza for the children of her son-in-law was preparedbefore it was dark by Matenzeni,)

/ / f /(32)a* Mombe idzi dzichadya chete nhasi mumunda uyu.(These cattle will surely lay waste today this field.)

b. Mumunda uyu muchadyrwa chete nliasi nemombe idzi,(This field will surely be laid waste today by these cattle.)

Nevertheless, the passive rule as it has been formulated does

not solve all the problems related to it. The rule would be more ade­quate if it could be so formulated that given the structures which under­lie the sentences in (9) and (10) and those in (2 3 ) and (24) it would

4. This sentence could also mean "At the cattle-pen were killed a snake and some children." In this discussion however this second reading is inconsequential.

185

be predicted that two passives are possible with the latter as shown in.(2 5) and (2 6), but that only one passive sentence is permitted with the former as shown in (13) and (1 4) respectively. As it is there does not seem to be a way of blocking the rule from producing the ill-formed sen-

tences in. (11) and (12). In the absence of a satisfactory \ my of speci­fying the correct deep structures which underlie sentences like those in (7) - (10) the passive rule as formulated on p. jgj . will suffice.

5.2 Pseudo-passive (cyclic)The application of the pseudo-passive rule is much the same

as that of the passive in 5*1. It will operate on a structure of the

/ / .5. It may be argued that by assuming that, for instance, Sara nzara m

(10) has the structure/" /Sara/ /nzara/ J and allowing the A-over-ANT? DTP NP NP NP NP

principle to apply, the ill-formed sentences in (15) end (16) are accoun­ted for* (For the A-over-A principle see Chomsky: 1964a; Poss: 1967a)This solution however is unsatisfactory for three reasons; (a) according

t /to this structure none of the noun phrases Sara and nzara is exhaustivelydominating the other, a condition necessary for the A-over-A principle,so that strictly speaking this principle cannot be said to apply here;(b) this structure does little, if anything at all, to capture the in-

/ /alienable relationship which exists between Sara and nzara; and (c)

given this compound noun phrase f ~ /Sara/ /mzara7 7 there will be aNT NP NP NP NP

difficulty passj_ve rule. Applying the passive then to the

structure underlying (10) will yield the ungrammatical sentence/” /Sara/ /nzara7_7 achekwa naFani*

NP NP NP NP NP NP

(*Sara finger-nails was cut by Fani.)tThis same sequence of words would be a grammatical sentence if Sara

and nzara are taken to be independent noun phrases, with Sara as subject NP, and nzara as object NP which has been front-shifted in this sentence*

386

following type:W?1 - AUX - V - FP2*

Like the passive it will interchange the subject NP with the object NP;

it will introduce the pseudo-passive formative /ik/ into the verb; and it will also introduce the agent formative /NA/.

However, the pseudo-passive, which is triggered off by an ele­ment in deep structure, differs significantly in meaning from the ordi­

nary passive* The meaning of the pseudo-passive is roughly that the deep structure object has the potentiality to undergo the action or state expressed by the simple verb. Note that the pseudo-passive process is very productive. The (b) sentences in the examples which follow haveundergone the pseudo-passive transformation,

f / /(33)a. Vanhu vanodya sadza.(People eat sadza.)

/ / / /b, Sadza rinodyika nevanhu.(Sadza is edible as far as people are concerned.)

(34)a* Basa rabaya varume ava.(Work has worn out these men.)

b. Varume ava vabayika nebasa.(These men have become worn out with work.)

The type of verb phrases which allow this rule to operate is the same as that which allows the simple passive to apply.

The pseudo-passive rule may be represented in formal terms asfollows:

Rule (2) Pseudo-passiveX - pseudo-pass. - NPX - AUX - V - (NP£) - NP^ - Y

1 2 3 4 3 6 7 8

1 0 7 4 5+ik 6 0 8 NA +~3 ^

condition.: 3 and 7 must not be referentially identical.Another difference with the simple passive rule is that, more often than

18?

not, after the pseudo-passive has operated, Agent Deletion applies. (For

the rule of Agent Deletion see subsection 5.3 below.) In the sentences which follow, though both the (b) and (c) sentences may be grammatical, the (c) sentences are more acceptable than the (b) sentences.^

(35)a* Maticha anonyora pabhodhi.(Teachers write on the blackboard.)

b. ?Pabhodhi panonyoreka nematicha*

(on the board - it can be written - by teachers = Blackboards can be written on by teachers.)

c. Pabhodhi panonyoreka*(on the board - it can be written on =Blackboards can be written on.)

(36)a. VaCharwe vanonyim/ VaSaru doro.(Charwe refuses to give Saru some beer.)

/ / / s / / / / /b. ?VaSaru vanonyimika doro naVaCharwe.(Saru can be refused some beer by Charwe.)

c. VaSaru vanonyimika do ro.

(Saru can be refused some beer.)/ / /(37)a. Vanhu vanzwa zvamataura*

(The people have understood what you said.)b. ?Zvamataura zvanzwika nevanhu.

(What you said has been understood by the people.)K * *c. Zvamataura zvanzwika.

(What you said has been understood.)The passive rule and the pseudo-passive rule are normally mutually

7exclusive in any simple sentence.

«*ly6. The notation ( ?) signifies that the sentence is marginally accept-

able to some Shona speakers.bub Is-reJ'ectod alte-get-hor by othoro.

7. I say "normally" because with a few verbs both the pseudo-passive

188

5.3 Agent Deletion (cyclic)The rule of Agent Deletion is meant to account for the absence

of the agent in sentences like the following:(3 8) Imba yedu yakapiswa makei,

(Our house was burnt last year*)/ i . /(39) Pahwahwa hapabvrke*

(at beer - not get away from = It is difficult to get away

from a place where there is beer.) it is quite clear that in (3 8) the grammatical subject is not the same as the logical subject* Somebody or something must have been the gram-

b£-and the simple passive formatives may^found occurring together in thatorder, e.g., -o^n-ek-w-a (be seen) cp -oila (see)

/ / /-naw-ik-w-a (be heard) cp -nswa (hear, under­stand)

-wan-ik-w-a (be found) cp -wana (find)Worse still, in some cases, the pseudo-passive, the reciprocal and thesimple passive formatives may occur together, e*g*/ / / / y

-ziv-ik-an-w-a (be known) cp -ziva (know)f / t /-d-ik-an-w-a (be loved, wanted) cp -da (love, want)

No doubt these verbs present some difficulty* Normally the presence of the formative /Ik/ (or /w/) in a sentence signifies that the pseudo­passive (or simple passive) has applied* Taking the verb -wana as an example, we see that it is usually followed by an object noun phrase,

Jeke anowana makuponi.(Jack gets (petrol) coupons,)

The underlying structure of this sentence meets the structural descrip­tion for^either the pseudo-passive only as in:

Makuponi anowanika naJeke*((Petrol) coupons can be obtained by Jack.)

or the simple passive as in:/ / /Makuponi anowaniwa naJeke,

((Petrol) coupons are obtained by Jack.)After the pseudo-passive has applied the structural description for the simple passive is no longer met, and vice versa. It would appear then that the occurrence of the formatives /ik/ and /w/ in anowanlkwa is not a result of the successive application of the pseudo-passive and the simple passive transformations. It follows, therefore, that, as trans­formational processes, the pseudo-passive and the simple passive are mutually exclusive in any simple sentence*

189

matical agent responsible for burning the house* But then this somebody

or something is not overtly expressed in the sentence* The same applies

to the sentence in (39)* It has got to be some person who can perform

the action of moving away from a beer party. In each of these two sen­

tences the agent has been left out. The rule which deletes the agent in

this way may be represented as shown below.

Rule (3) Agent Deletion X - na + NP - Y

1 2 3>

1 0 3 7It must be emphasized though that the phrase na + NP ought

to be one which has resulted from a previous application of either the

simple passive or the pseudo-passive transformation. Notice that in

Shona na * NP can signify other ideas also that have nothing to do with

agency at all. For instance,

na + NP may indicate instrumentality as in (40);

(40)a. Baba vauraya mombe nedemo.

(Father killed a cow with an axe.)i / /b. Mombe yaurayiwa nedemo nababa.

(A cow was. killed with an axe by father.)

c, Mombe yaurayiwa nedemo.

(A cow was killed with an axe.)

Agent deletion cannot apply to (40a) to delete the phrase ned e m o . It

only applies to (40b) and the result is (40c). Note however that strictly

speaking agent deletion only applies to unspecified agents (e.g. munhu

(a person), chinhu (a thing), zvinhu (things)), not babsf, because ofQ

the condition on the recoverability of deleted material.

8. For some discussion on the recoverability of deleted elements see

J.J.Katz and P.M.Postal in An Integrated T h eory of Linguistic Description

190

(b) Co-ordination is another idea which is expressed by na + NP,

(lfl)a. Tafa naPeni7 vakarova ticha gore rakapera.

(Tafa and Peni beat up a teacher last year.)

b. Ticha akarohwa gore7 rakapera naTafa7 naPeni.7

CA teacher was beaten up last year by Tafa and Peni*)f / / / /e, Ticha akarohwa gore rakapera,

(A teacher was beaten up last y e a r , )

Because Tafa^naPend.7 constitutes a conjoined structure Agent Deletion

cannot apply to it to delete naPeni without destroying the conjoined

structure. This rule can, however, apply to (tlb) to delete n a T a f / n a P e n i 7

and resulting sentence is (hlc).

Agent deletion, is ordered after both the simple passive and

the pseudo-passive rules.

Reflexive (cyclic)

In Shona, just as in many other languages, whenever the object

noun phrase is identical with the subject noun phrase and the two are

in the same simple sentence, the former constituent is obligatorily re-

flexivized as the sentences below illustrate.

(AB)3-* */Mombe iye/-^ yakuvadza /mombe iye/^

(*/That cow/^ hurt /that cow71*) b. Mombe iye y a zvikuvadza{

(That cow hurt itself.)

C^3)a. *£Mwana/1 acharuma /mwana/-^ /rurimi/^.

(*/The child/^ will bite /the chil<|7-^ /the tongue72*

b. Mwana7 a^hazviruma rurirai7

(The child will, bite itself on the tongue,)

(196A) p*A3ffF-ftd- p.80ff; Chomsky in Aspects of the Theory of Syntax p.IMfffj K.Ruwet in An Introduction to Generative Grammar (1973) (in

translation) p. 218 -- 219*

191

This rule may he represented formally as follows;

Rule (if) Reflexive

X - NP, - A U X - V - N P - YX1 2 3 4 5 6

1 2 3 p 5 H H 0 6l+reflex,

Now NP

g|NP I will always be spelt out as zvi (or dzi in some forms ofL+reflex^J

the Manyika dialect) irrespective of the class to which the head noun of the noun phrase belongs.

The word "identity” requires comment. Just as during the dis­cussion of the passive rule partial co-referentiality between the subject

NP and the object NP was enough to block the sentence from passivizing, so also will partial co-referentiality block reflexivization. For an object NP to be reflexivized it must be strictly referentially identical with the subject NP in that sentence. To this end consider again (16a)

which is repeated here for convenience as (44) •, i f / / / r /(44) Sara anorova vana vake.

If the possessive pronoun -ke in this sentence refers to Sara*, then the structure which underlies this sentence is roughly that in (45)*

(45) Sara anorova /vana / S a r a ane vana/ 7*NP S S NP

(Sara beats up /children ZPara ^as children/ _/*)NP S S NP

The possessive construction here is assumed to derive from an embedded- / - /sentence. It is: clear that the object NP in (45) > namely, /vana /Sara

NP S

ane vana7 J is not altogether referentially identical with the subject S NP

/NP, namely, Sara. In this case reflexivization blocks as shown in (46),/ f / /(46) *Sara anozvirova.

The sentence in (4 6) cannot be said to be the reflexive sentence of (44)*

192

This does not however imply that (46) is ungrammatical per se„ It canbe grammatical but only if it is deemed to derive from (4?)*

/ / / / /(4?) Sara^ anorova Sara^(Sara^ beats Sara^*)

This sentence, that is (V?), meets the structural description for the

reflexive rule, but not the sentence in (45)»

5*5 Mood Incorporation

Complementizers have been taken to consist each of two elements, namely, a complementizer formative and a mood, e.g..

(4 8)a. kuti 1+ indicative/ b* kana / + participial/

(For the full list of complementizers see subsection 2.10.) These comple­mentizers serve to introduce embedded sentences as has already been stated above. The embedded sentence which is introduced by complementi­zers has the following tree structure;

(49) o

COMP $

KTP AITX VP

The symbol B signifies that this is an embedded sentence, while the symbol S signifies; a sentence similar in all relevant respects to the S which is taken as the initial symbol of the grammar* The rough diagra­mmatic representation of the deep structure of the sentence in (5 0), which contains an embedded sentence, is found in (51)•

(50) Rudo^ anoziva kuti vana vanoshereketa*(Rudo knows that children are mischievous.)

193

(5D 1

Rudo V NP

ziva N S

f+N H COMP S| +CL8 fl+pro_J I

_kuti^•i-indic./

NP VP

vana V

shereketa

Xu any sentence in. Shona the nucleus of the predicate, hy which is meant in this case the subject concord, the tense and the verb proper as form­ing one word, must appear in one or another of the moods that exist in

the language* Assuming that the diagram in (51) is correct, we notice that mood is dominated by the node COMP which lies outside S^* In the surface structure of this sentence it is not the complementizer formative kuti, but rather the nucleus of the predicate of S2 , which carries the indicative mood* Nevertheless, it is kuti which determines that the predicate of should be in the indicative mood* If another complemen­tizer formative is substituted for kuti, e.g. kana, the predicate of S2 will be in the participial or consecutive mood, but certainly not in

the indicative mood.

such sentences as the one in (50), it is proposed to set up a rule which will move the feature mood from under the domination of COMP and place it somewhere in the auxiliary, maybe as $ sister node of Tense, (Tense is suggested here because, given a mood, there will be variations

of tone patterns within that mood which are occasioned by Tense.) Let

In order to characterize what happens in the derivation of

19 4

us label this rule the Mood Incorporation rule. It may be formalized

as follows:

Rule (1) Mood Incorporation

X - ^complementizer— ' /rnood/ - /NP - A U X - Y7 /Z S formative S S B

$

1 2 3 ^ 5 6

1 2 0 1 3 + 9 6

This rule is meant to capture the idea that the embedded sentence will

be in a mood the choice of which is determined by the complementizer.

3.6 Complementizer Placement (cyclic)After the Mood Incorporation rule has applied there is need

for another rule which will take the complementizer formatives from

their position before the subject NP to a position immediately after

this subject NP. That such a rule is desirable is evident from an exami­

nation. of the following pairs of sentences.t / t t f /(32)a. Pita anofunga kuti Kondo anodzoka raangwana.A. * , tb. Pita anofunga Kondo kuti anodzoka mangwana,

(Peter thinks that Kondo will return tomorrow.), . / / / / / / / / /(53)a* Fani akafembera kuti sekuru vanosvika nhasi.

/ / / / / / / /b. Fani akafembera sekuru kuti vanosvika nhasi,

(Fani calculated that uncle would arrive today.)(54)a* Rwodzi^ akanganwa kana Motsi^akataura mazv/i iwayo.

b„ Rwodzi^akanganwa Motsi^kana akataura mazwi"iwayo,

(Rwodzi has forgotten whether Motsi said those words.)t f t / /(55)a. Kana Roda akapasa zvingashamise.

b. Roda^ kana akapasa zvingashamisel

(If Roda passes it will be surprising.)' ' ' ' t . A / A*Zva mvura yapera mutsime zvichanetsa.

(since - water - it has finished - in well - it will be hard)

195

b* Mvura zvayapera mutsime zvichanetsa,(water - since it is finished - in well - it will he hard = Since there is now no water in the well there will he

a problem.)(57)a* *Ku Matinyadza chaiye anosvika pano zvinoshura.

(for to - Matinyadza - she arrives - here - is an ill-omen)b. Matinyadza chaiye kusvika pano* zvinoshura.

(For Matinyadza to come here is an ill-omen (for us)0

Both sentences in each of the pairs in (52) - (55) well-formed, but only the (b) sentences;in (56) and (57) are grammatical. The sentences in each pair have the same semantic content. Note that in the (a) sent­ences the complementizer formatives lie outside the embedded sentence,

while in the (b) sentences they occupy a position immediately after the subject NP. This is the only structural difference between the sentences in each pair. It can be seen from these sentences that while the placing of the complementizer formatives after the subject NP of the embedded

sentence is optional in (5 2) - (55)? it is obligatory in (56) and (57)*I shall refer to this process as complementizer placement.

Rule (6) Complementizer Placement

X(kuti"i /kana , \ ku

^NP - AUX

2

0

3

3+2

y 7~s

obligatory with ku and zva other- wise it is

optional

Note that this rule is considered to apply only after Mood Incorporation has applied.

It might be suggested that the (b) sentences in (52) - (55)

show instances of object raising (discussed in 5*7 below) rather than complementizer placement. That such a suggestion is fallacious is easily

196

demonstrable. Let us consider (52b) as an example.. If Kondo occurs before the complementizer formative kuti by virtue of the rule of object rais­

ing, then it follows that it is now a direct object NP of the verb -f unga for this is what the rule of object raising does. That being so, the object anaphora and the passive rules, for instance, should be able to apply, but they do not, as the ungrammaticality of (58a) and (58b) indi­cate respectively.

(58)a* *Pita anomufunga kuti anodzoka mangwana*/ / / / / /b* *Kondo anofungwa naPita kuti anodzoka mangwana.

These same transformational rules will yield negative results in respect

of the (b) sentences in (53) a&d (54) shown in (59) and (60) respec­tively.

(59)a. 4fFanj/ akavafembera kuti vanosvika nhasi.b. *Sekuru vakafemberwa naFani^kuti vanosvika nhasi.

. x // / / / / / / J J(60)a* *Rwodzi amukanganwa kana akataura mazwx iwayo.i t t / /, / < fb 8. *Motsi akanganwiwa naRwodzi kana akataura mazwi iwayo#

In (55"b) object raising could not have applied by virtue of the position

occupied by the complement clause in this sentence. Since this is anembedded sentence functioning as a subject it might be suggested thatsubject raising (see 5^8) has occurred instead. This suggestion too iseasily discounted because if subject raising has occurred in (55^)j then

/Roda is now the grammatical subject of the containing sentence and con­sequently it should now control the concordial agreement in this higher

f f t /sentence, but it does not* The concord zvi in zvingashamise is neither dictated nor controlled by the class prefix of Roda. If Rode/ controlled the concord in the higher sentence we would have the following sentence which however is also grammatical.

(61) Roda kana akapasa angashamiseC(If Roda passes she could surprise (us).)

/It is quite clear therefore that Roda occurs before kana in (55b), not

197

through, the rule of subject raising, but by some other process* It is

claimed here that this change in word order is occasioned by the process

which I have called complementizer placement*Here is a further illustration of the operation of this rule.

The deep structure representation of the sentence in (56) would be some­

thing like the tree in (62)*

(6 2)

COMP

VPNP

NPV

pera

netsa

mutsime

(63)

+proNP COMP

NPmvura zva

pera mutsime

netsa

After applying Mood Incorporation and then Complementizer Placement to the structure in (62) the result is the tree in (63)* complementizer

198

formative zva now occurs after the noun phrase mvura* Further examples showing how the rule of complementizer placement operates can be seen

in the sentences in (62) - (65) in Chapter Four*

5*7 Gender Copying (cyclic)Each of the following strings of words is a noun phrase*

(6A-)a, chipunu chose (the whole spoon)

b* zvipunu zvose (all the spoons)c* tupadza twose (all the little hoes)d* kamba kose (the whole little house)

In each noun phrase there is concordial agreement between the head noun

and the quantifier -ose (all, whole)* The concords are underlined in these noun phrases. If the concords on the quantifier are mixed up the resulting strings will be ungrammatical noun phrases as is shown in (65)*

(65)a . *chipunu z v o3e b * * chipunu two's e

- f 4- ' *c» wzvrpunu twose/ / /d * * kamba clip s e

When making up lexical entries for the quantifiers, and any other nounqualifiers for that matter, it is not necessary, and indeed it would begrossly uneconomical, to list the quantifier together with the quantifierconcords as in (66).

/ /(66)a* chose (all, whole)b* twose (all, whole)

The quantifier concords are predictable once we know the gender or class to which the head noun belongs* Thus, if we are given a head noun with

CL ±J as one. of its cluster of features we can predict without any difficulty that the quantifier is also going to have obligatorily the

-7 /feature £+ CL^_/* When the. noun, chipunu appears in the lexicon it will include among its features the following:

199

(67) -punu

+count

spoon/ /while the quantifier chose will appear something like this:

(68) ~ / /-ose-^Quantifier

all, whole

The noun phrase chipunu chose would then he derived by invoking an agree­ment transformational rule which copies the gender or class feature of the head noun on to the quantifier. Later rules, including the noun pre­

fix segment rule which will be discussed briefly in 5.16, will then applyto give us the noun phrase chipunu chose.

Concordial agreement is not only found within a noun phrase,but also between the subject noun phrase and the predicate.

/( 6 9 ) Chingwa chinonaka.(Bread tastes nice.)

b*. Kamba aka kanotonhora.

(This little house is cold.)t / f i , , /c. Varume yanonwa- hwahwa*

(Men drink beer.)As before, if the concords are interchanged the results will be ungram­

matical sentences as in (7 0).(70)a , *Chingwa kanonaka.

b. *Chingwa vanonaka.c. *Varume chinonwa hwahwal

It is observed here that the feature / + CL. / of the head noun of the

subject noun phrase is copied on to the predicate in order to effect

200

concordial agreement.

The formulation of the agreement rule, or gender copying asit will he referred to in this study, in Bantu languages is a subject

gof debate in current linguistic studies* One solution is to set up twoseparate rules, one to cope with agreement within a noun phrase, and the

other to account for agreement between the subject noun phrase and thepredicate. Such rules would be rtttfghly as follows:

(7 1)a. agreement within a noun phrasef Ad j *J Demonstr,

< Selector 1 Quantit.

— — y^Possessiye

2 3 k

X+N+CL,

l+CL,

b* agreement between the subject NP and the predicate'TN -1

X * CL,X*

- Y - AUX - Z

NP#

NP

1 2 3 k --- 1 , 2, 3+CL.

l±

However desirable these two rules may seem to be, they make a misleading claim, namely, that the L+ CL±J feature is copied on to the qualifier as one operation and on to the predicate as another, independent opera­tion. Our grammar would be more adequate if this feature-copying opera-

9. For more comprehensive discussions on concordial agreement in Bantu languages in the transformational model see, among others, E.A.Gregersen (1967) in "Prefix and Pronoun in Bantu", Memoir 21, I.J.A.L.; Talmy Givon in "Studies in ChiBemba and Bantu Grammar" in Studies in African Lingui­stics » vol. 3 5 Supplement 3, June 1972; Frank Heny in "Bantu Lexical

Classes and Semantic Universals" also in Studies in African Linguisticss

vol. 3, Number 2, August 1972.

201

tion were performed simultaneously on both the qualifiers and the predi­cate.

Givon has proposed a rule which he put in the form of a schemaand which goes quite a long way towards resolving concordial agreement

10 iin Bantu languages. Following Givon let us first define the categories of lexical item which are concordable in Shona. Let X he such a category.

X ________^ (Adjective) (Demonstrative) (Selector) (Quantifier)(Possessive) (AUX)

Concordial features will then he spread on to all such items in the re­

levant positions. The rule then works like this:

Rule (7) Gender Copying

subject

This rule is to be interpreted as follows:

(a) agreement between noun and qualifier (s)

HP

X N X

stem+CL.

The constituents to which the class affix is copied may occupy a posit­ion before or after the controlling noun as illustrated above. (For examples see the noun phrases in (6i|-) above.)

10. op. cit, p.75 - 109

202

(b) agreement between the subject NP and the predicate

AdjectiveDemonstrativeSelectorPossessiveQuantifier

e.g. (i) Chingwa chaiva chishoma(The bread was not enough.)

(ii) Tuhukwana tunokwana tuna.(The chickens add up to four.)

/ / / / / /(lit) Zvinhu zvichava z vein wan ak o m an a wake.

(things “ will be - of son - his aThe property will be inherited by his son.)

The first case of this rule schema spreads the class feature on to every

constituent in the noun phrase in which a given head noun occurs. How­ever I have not found a genuine case of the application of this subpart in Shona. The view adopted in. this study is that all the concordable categories are somewhere in the predicate at the time that this ruleoperates. These categories will therefore acquire the class feature

from the application of the second subpart of this rule.The foregoing discussion does not pretend to have described

exhaustively the phenomenon of gender copying in Shona. Far from it. Assumptions were made without justifying them. For the purpose of this study, however, it was sufficient to show in broad outline how gender copying operates in Shona without going into minute details.

5.8 Object Raising (cyclic)The next rule to be discussed is one whose existence is not

so apparent from a casual examination of surface structures as are the

203

others which we have seen* This transformation helps to explain the synonymy in the following sets of sentences:

(72)a* Ticha anoda kuti vana vatambe bhora.(The teacher wants the children to play football.)/ / / / / t f l tb. Ticha anoda vana kuti vatambe bhora*

(The teacher wants the children to play football.)f I t i f f / / /c. Vana vanodiwa naticha kuti vatambe bhora.

(The children are required by the teacher to play football.)(73)a. Vanhu vazhinji^ vanoziva kuti Ruzvidzo anodada.

(Many people know that Ruzvidzo is conceited.)/ / / / / / /b, Vanhu vazhinji vanoziva Ruzvidzo kuti anodada.

(Many people know Ruzvidzo to be conceited.)/ / / / / t ( t jc, Ruzvidzo anozivikanwa nevanhu vazhinji kuti anodada.

(Ruzvidzo is known by many people to be conceited.)(7k)sw Zuze achaona kana basa racho rakanaka,

(Zuze will see whether the job is suitable.)/ / / / / / / / /b. Zuze achaona basa racho kana rakanaka*

(Zuze will see whether the job is suitable.)/ / / / / / / / /c. Basa racho richaonekwa naZuze kana rakanaka.

(The job will be seen by Zuze whether it is suitable.)Since in each set the lexical items in each sentence are the same and the meaning is also the same we would expect these sentences (i.e. the sentences in each set) to have identical deep structures* In (72) for instance, (72b) differs from (72a) in that the noun phrase vana occurs before the complementizer formative kuti; and (72c) differs from (72b) in that it is in the passive* While it is relatively easy to establish the relationship between (72b) and (72c), that is, one is active and the other passive, it is not as easy to relate (?2a) and (72b); yet itis the relationship between these two sentences which is crucial here.

Sentence (72a) clearly has as its object noun phrase a coraple-

20 k

f i / t /Brent clause, namelys kuti vana vatambe bhora. That this clause is func­tioning as a noun phrase is easily demonstrated. This complement clause passes the tests for object noun phrases as shown below*

(75)®-* pseudo-cleft;Kuti vana vatambe bhora ndizvo zvaanoda ticha,(That the children play football is what the teacher wants.)

b, object anaphora;/ / / / , / / / / / .Ticha anozvida. (i.e.. kuti vana vatambe bhora)

(The teacher wants it, (i.e. that the children play football.)c, interrogative pro-form;

/ / / / , / / / / / .Ticha anoda chii? (reply: kuti vana vatambe bhora)(The teacher wants what? (reply: that the children play

football)d, passive;

/ / / / / / / / ,/Kuti vana vatambe bhora zvinodiwa naticha.(That the children play football is wanted by the teacher.)

Now if (72b) shares the same deep structure with (72a) it must be the case that it too contains a sentence as its object noun phrase. But here our tests produce negative results,

(76)a. pseudo-cleft;"Kuti vatamba bhora ndizvo zvaanoda vans' ticha.

b* object anaphora;*Ticha anozvida vana. (i.e. kuti vatambe bhora)

c. interrogative pro-form;*Ticha anoda vana chii? (reply: kuti vatambe/ bhora)

d. passive;/■"Kuti va

The ungrammaticality of the sentences in (76) shows that the stringf i t J!kuti vatambe bhora is not a noun phrase in (72b). This observation

seemingly reduces the chances of demonstrating that both (72a) and (72b)

t f t i /. ( / f ,*Kuti vatambe bhora zvmodiwa vana naticha.

205

have the same deep structure.In order to show that these two sentences have the same deep

structure, it must he shown that some process has operated on the noun phrase complement in (72h) which destroys its structure as a noun phrase and which creates a new structure which is incapable of passing the tests for object noun phrases* The crucial clue here is to be found in (72c), namely, the existence of the passive transformation* Let us briefly look at the passive transformation again* The structure to which it applies must meet two basic conditions. First, there should be a subject noun phrase and an object noun phrase and these two must not be referentially identical* It is these two noun phrases that the rule will interchange. Second, and for the moment more important, it is necessary for these two noun phrases to be in the same simple sentence as illust­rated in the sentences below.

(77)a* Vanhu vanoziva Kenl*(People know Keni.)

( / / i / /b» Keni anozivikanwa nevanhu.(Keni is known by people.)

(78)a. Vanhu vanoziva kuti Ruzvidzo akaponda Ken/.(People know that Ruzvidzo murdered Keni.)

i f f f / / / / /b. Kuti Ruzvidzo akaponda Keni zvinozivikanwa nevanhu*(That Ruzvidzo murdered Keni is known by people.)

/ / / / / . / / / /c. *Keni anozivikanwa nevanhu kuti Ruzvidzo akaponda.(*Kerti is known by people that Ruzvidzo murdered.)

/ / / / / / / /d. Vanhu vanoziva kuti Keni akapondwa naRuzvidzo.(People know that Keni was murdered by Ruzvidzo.)

Let us compare the rough deep structures of the sentences in (7?b) and (?8d) which appear below as (79) and (80) respectively.

206

(79)

PreS

passive vanhu ziva Keni

(80)

vanhu V

ziva

+N+CL8+pro

COMP

VPPreS NPindie*/

passive Ruzvidzo V NPi iponda Keni

In (79) the two noun phrases- which are to he interchanged by the passivB are in the same simple sentence* This equally applies to the noun phra- ses Ruzvidzo and Kenlf in (80) : the passive can apply and interchange them because they are dominated by the same S, namely, Sentence

t(78c) is ill-formed because the noun phrase Keni, dominated by S^, is made to interchange with vanhu which is dominated by another S-node, namely, S^, The same applies also to Ruzvidzo in (80): it cannot be interchanged with vanhu while it (i.e* Ruzvidzo) is still under the domination of The passive rule cannot apply in such a case*

In order to show that (72a) and (72b) share the same deep structure, the applicability of the passive transformation in (72c)requires- some explanation* In particular it must be described how vana,

207

if it originates in the complement clause, comes to be dominated by the higher S® At this stage let us examine the deep structure representation of (72a)® This ‘is provided in (8l).

none of the noun phrases in either pair can be interchanged by the appli­cation of the passive since none is in the same simple S. But in (72c) at some point in the derivation vana must be in the higher sentence otherwise the passive would not apply. The deep structure of (72c) ap­pears below.

(81)

Ticha

da N S

r+N *1 cot| 4 - C L 8 I J1+pro I I

tamba bhora

/ / / /Although ticha and vana, and also ticha and bhora, are not co-referential

(82) 1

passive ticha V NP

da . = N S

COMP

tamba bhora

208

In order’ to resolve this difficulty it is proposed to set up a rule to11be called Object Raising, This seems to do two things. First, it deta­

ches the noun phrase complement which has one of the following comple­mentizers: kuti /+ subjunctive/, kuti /+ indicative/, kana /+ participial/,from under the NP~node which dominates it and attaches it under the domi-

1 Pnation of VP of the next higher S, Applied to the structure in (82) this first stage results in the tree structure in (83)•

(83)

PreS

passive ticha V

da N

+N+cl8+PFO

VPNP AUX/-f sub j,_/

vana V NPTm

no tamba bhora

VPNPPreS

passive ticha V

COMPda

VPJfcuti __ AUX/+ subj^/ j

no bhora11, cp Mit,t replacement- (Rosenbaum: 196?a); it-»substitution (R.Lakoff;

1968); raising rule (Grinder and Elgin: 1973); raising object (Stockwell et al : 1973)•

12, No cases involving this rule have been attested with the other complementizers.

209

In the second stage this rule substitutes the pronoun head +cl8+pro

of

the noun phrase which contained the transported complement clause withthe subject noun phrase of the complement sentence. This results in thestructure in (8if) above* If we did not have passive in the deep structure,the structure in (8-t) would then lead to the derivation of (72b).

But now we can consider the passive in the higher sentence in(72e). What we observe here is that the object raising transformationhas created a structure in the higher S to which the passive can be

/ /appliedj interchanging the noun phrases ticha and vana. This is nowpossible because the subject of the complement clause has been movedinto the higher S, and the condition for the passive rule, namely, thatthe interchanging noun phrases should be in the same simple 8, is nowmet. The substituted noun phrase becomes syntactically independent of the embedded sentence to which it belonged in the deep structure, andbehaves as if it were the direct object of the higher verb. It is forthis reason that the sentences of (76) are ill-formed. The direct object

Iis now vana and not the embedded clause. The embedded clause ceases to be the object noun phrase.

In this case the passive applies after object raising has applied. If in the embedded clause the passive had been present, there would have been a situation in which the passive applied first, then object raising, and then passive again (but in the second cycle this time). Object raising itself is ordered after gender copying otherwise S„ would be left without a subject concord and this would render thec.

sentence ungrammatical. Applying the passive* to the structure in (8 ) the resulting structure is (85)-

210

(85)

vana

COMP

AUX VP

V

no tamba

na ticha

bhora

This rule may be formalized as follows: Rule (8) Object Raising

X V -Z+°bJ._rais./

NP

+N+CL8 Jkuti + indie . 1 r~ i+pro Jkuti + subj. ► [(PreS) - NP - Y)_ — y^ana + par tic.. s s_

3 s__NP

1

1

2

2+5

3

0$

The constituent 0 3 in the structural change is now attachedS ' S

as a daughter node to the VP of the higher S.By setting up the rule of object raising we are able to account

for a number of facts in respect of the sets of sentences in (72) - (7A)*First, this analysis explains that the (a) and (b) sentences in thesesets have the same meaning because they have the same deep structure.The only difference between the )c) sentences and the others is that theformer, but not the latter, have the passive specified in their deepdeep structure. Secondly, this analysis explains passivization in thehigher S in the (c) sentences by showing how the subject of the embedded

211sentence is brought into the higher sentence so that it is in the same simple sentence as the noun phrase with which it interchanges positions.. Thirdly, it explains why the sentences in (75) are well-formed while those in (?6) are not..

Here are further examples to show how the rule of objectraising operates..

/ / / / / / / / / /(86) a* Taka akacherechedza kuti Dhongeri yakanga isipo..t / / / / i i t i tb.. Taka akacherechedza Dhongeri kuti yakanga isipo,

(Taka noticed that Dhongeri was not there.)I t / t t t i l l f tc,. Dhongeri yakacherechedzwa naTaka kuti yakanga isipo.

(Dhongeri was noticed by Taka that it was not there.). i t / / ✓(o?)a. Takawira anofungidzira kuti Mombe ndiye akaba mari.

i t / / /b. Takawira anofungidzira Mombe kuti ndiye akaba mari.(Takawira suspects that it was Mombe who stole the money.)

t t t f /c. Mombe anofungidzirwa kuti ndiye akaba mari naTakawira.(Mombe is suspected to have stolen the money by Takawira.)

(88) a. Vamwene vakao^na kuti muroora ane usimbe.f i t t t // / 'b.. Vamwene vakaona muroora kuti ane usimbe,

(mother-in-law - saw - that - daughter-in-law - was lazy =The mother-in-law saw that the daughter-in-law was lazy.)

/ / / / / / / . / / i fc:. Muroora akaonekwa kuti ane usimbe navamwene,(The daughter-in-law was seen to be lazy by the mother-in- law. )

t / t i f t(89)a. Vanhu vazhinji vanotendera kuti Mwari ariko.t * i t t 1 fb.. Vanhu vazhinji vano tender a Mwari kuti ariko..

(Many people believe that God exists.)i f i f t f ta. Mwari anotenderwa kuti ariko nevanhu vazhinji,

(God is believed to exist by many people.)

5*9 Subject Raising (cyclic)While object raising replaces the pronoun head of an object

noun phrase by the subject noun phrase oi£' the complement clause, there

212

is another similar operation which is illustrated in the sentences in (90) - (93).

/ / / / / ✓ / / ✓(90)a* Nyava iyi zvayadai zvinoreva kuti totoenda kudare.(Since this case has taken this turn it means that we have to go to court,)

Nyaya iyi inore^va kuti totoenda kudare zvayadaxl (This case forces us to go to court since it has taken this turn,)

(91)a, Kuti Podzorimwa adye sadsa nemunwo zvmoratidza kuti angat /ava nenzara,(For Podzorimwa to take his sadza with vegetables indicates that he was hungry.)

Podzo'rimwa anoratidza kuti anga ava nenzara kuti adye sadza/ / /nemunwo»

(Podzorimwa showed that he was hungry by taking his sadza with vegetables,)

(92)a. Zvaita kuti vatezvara vauya.(It is a good thing that the father-in-law has come.)/ / t ib* Vatezvara vaita kuti vauya,

(The father-in-law has done a good thing by corning.)(93) a. Ivte zvawatora marx^ yake zvichakukonzera kutaura.

(Talcing his money as you have done will cause you some trouble,)

b. Iwe/ uchazvikonzera kutaura zvawatora mari^ yake!,(You will cause yourself some trouble by taking his money as you have done.)

The sentences in each pair have the same meaning. In the (a) sentences above the underlined noun phrases are the subject noun phrases of the embedded sentences. These same noun phrases5 by substituting for the pronoun heads have become the subject noun phrases of the higher S, In

213

other words the subject noun phrases of the embedded sentences in the(a) sentences have been raised to become the subjects of the higher sentences in the (b) sentences* We shall call this operation SubjectRaising. This rule may be represented as follows:

Rule (9) Subject Raising — —

X -

NP

kuti 4-n.ndic, kuti 4-subJ. ,zva 4-par tic,

X/“S

4

AUX „ X,/4-rais/

NP

Let us now consider how this rule applies in the derivation of (92b), The deep structure representation of the sentences in (92) is provided in (94)*

(94)

COMP4-pro

kutiz." indie./

vatezvara Tns

a

V

uya

it a

With the verb -ita a complement clause in subject position seems to be extraposed obligatorily to the end of the sentence. Applying the rule of extraposition (see 3*11 for a treatment of this rule) to the struct­ure in (94) the result is the tree in (95)* From this structure is eventually derived (92a).

21.4

(95)

VP s

COMP

ita _kuti _ I* indie./

NP AUX

vatezvara Tns

3. uya

If to the structure in (94) is applied the subject raising rule the re­sultant structure is (96) below,

(96)

vatezvara

kuti VPAUX£+ indie *y

Tns

cl

v

uya

ita

As was the case in (85) s because the verb in the higher S is -ita, extraposition applies and this results in (97)*

215

(97)

COMPVatezvara

ita _ kuti __ AUX indie */

Tns

a uya

After applying later rules (92b) is derived.But extraposition after subject raising is on the whole option­

al as is evidenced by the fact that alongside (90b), (91b) and (93b)there are also such sentences as (98), (99) and (100) respectively inwhich the embedded sentences in question are underlined.

(98) Nya5ra iyis zvayadai^, in-or/va kuti totoenda kudare,(^his cases since it has taken this turn, forces us to go to court*)

(99) Podzorimwa, kuti adye sadza nemuriwo, anoratidza kuti anga/ /ava nenzara.(For Podzorimwa to have his sadza with vegetables shows that he was hungry*)

(100) Iwe, zvawatora mari^ yake , uchazvikonzera kutaura.(By taking his money as you have done you will cause yourself some trouble.)

In these sentences subject raising has applied but not extraposition.A comparison of the sentences in (93a) and (93b) reveals some­

thing which strengthens the view that the rule of subject raising exists in Shona* In (93a) the subject, namely, the pronoun head, and the object, namely, lyie (represented here by the object prefix ku) of the main verb -konzera are different* In (93b) however the subject is now iwe but the

216

object is still iw/» These two noun phrases are now clause-nrates, hence reflexivization operates and this is represented by zvi in the sentence* It is quite clear that the subject of the embedded sentence, i.e. iwef, has been raised to become the subject of the containing sentence,, thus creating the structure for the reflexive rule to apply«

The rule of subject raising helps to explain why the pairs of sentences in (90) - (93) have the same meaning in spite of their diffe­rence in surface structure*

5.10 Equi-NP Deletion (cyclic)The application of this rule is responsible for the contrast

which exists in the following sentences:(101) Tukomana utv/u tunoda kuti zvipfuwo zvxdye pizi dzartgu.

(These small boys want the domestic animals to destroy (lit.eat) my pea crop,)

(102) Tukomana utwu tunoda kuti tudye pizi dzangu.(These small boys want to eat my peas.)

(103) Zvipfuwo zvinod/ kuti zvidye pizi dzangu.(The domestic animals want to■destroy my pea crop.)

/ /In (101) the subject of the embedded sentence, namely.zvipfuwo, is pre-sent in the surface structure. But in (102) and (103) the subjects ofthe verbs tudye and zvictye are not overtly expressed. Nevertheless thesubjects of these predicates can be supplied from the higher sentence.

f i i fIt is quite evident that these subjects are tukomana utwu and zvipfuworespectively. What has happened in (102) and (103) be explained inthis way: the subjects of these embedded sentences were present in thedeep structure but were subsequently deleted by a transformational rule.The result is that they do not appear in the surface structure.

It is possible to adduce syntactic evidence to show that,for instance, tukomana utwu is present at some point in the derivation

/ Vof (102). This is clearly shown by the concord tu in tudye which is

/ / / /evidently a reflex of til in tukomana. In zvidye the concord zvi is also clearly a reflex of the zvi in zvipfuwo^ Whereas (102) and (103) are grammatical 9 (10A) and (105) are not, precisely because the concord zvi in (104) does not agree with the noun prefix tu in tukomana<, and the

t /concord tu in (105) does not agree with the noun prefix zvi in/ / / t / t / / /(10if) *Tukomana utwu tunoda kuti zvidye pizi dzangu.

/ / t i t / / / /(105) *Zvipfuwo zvinoda kuti tudye pizi dzangu*(These sentences are however grammatical if the concords zvi and tu inthe embedded sentences refer to some noun phrases already mentioned elsewhere in the discourse. The noun phrases to which they refer willhave been deleted by Equi-NP deletion in any case.) In (102) the nounphrase tukomana utwu must have been present In the embedded sentence at some point in the derivation so that gender copying could operate. The same applies to the noun phrase zvipfuwo in (103).

The rule which deletes the subject of the embedded sentence, commonly referred to in the literature as Equi-NP Deletion (Equivalent noun phrase deletion), applies only when the subject of the lower sent­ence is identical with some noun phrase in the containing sentence, so that the former, when deleted, can be supplied* This rule may be stated as follows:

Equi-NP DeletionRule (10)

X - (NP) X

1

1

fkana') mJkutil /(PreS) jku C SW a J

NP - AUX VsX -(NP) - X

condition: 5 = 2or 5 = 8

where 2 or 8 is the NP nearest to 5 in the S which is being proces

218

As an example of how this rule operates consider the derivationof the sentence in (106) whose underlying structure is given in (107).

/ / / / / / / / /(106) Vakomana vanoziva kuti vakatadza.(The boys know that they made a mistake.)

(107)

vakomana

I +CL8 I1 +pro I

COMP S

VPNP AUX/+ indie./

vakomana Tns

Iaka

V

tadza

Equi-NP deletion can apply to this structure (taking its proper place in the cycle) since the subjects of the two sentencess namely and S^are identical. This results in the tree structure in (108).

(108) S-

vakomana V

COMP+N +CL8 +x>ro

kuti

aka tadza

Later rules will then convert (108) irito the superficial form in (106)

219

5.11 Extrapo sition (non-cyclic)Consider the structural relationship that exists: between the-

feentences in (109) below.

That these two sentences are semantically equivalent is: uncontroversial. The main difference between thenr is structural, that is, in (109a) the complement clause occurs; before the main verb while in (109b) it occurs after the main verb. The deep structure of (109a) is (110a) which is highly abbreviated.

t i l t / / /(109) a., Kuti mwana aite nungo zvakaipa.(For at child to be. lazy is bad.)

/ / / / / /, , / /b.. Zvakaipa. kuti mwana aite nungo..(It is bad. for a child to be lazy.)

(110)a. S.1

VP

K S V

COMP

_ kuti _ /+ sub^*/

mwana

ipa

ita nungo

The diagram, which represents the sentence in (I.09b) is (110b.) below.

220

(llO)b., S.1VP s

/COMP S

ipa __ kuti w 1+ sub3^7j xpro I

mwana

ita nungo

It is quite common in Shona for the complement clause of a subject, nounphrase to be optionally transported, or extraposed, to the extreme rightend of the whole sentence.. This extraposition of the complement clausemight appear to be the same as. Subject Postposing (this rule is discus­sed in 5*i7 below). But it is not the same process;. To this end considerthe following sentences:

(111)a* *Kuti vatezvara vauya zvaxta.

As was noted in subsection 5*9» with 1: -ita as the main verb the comple­ment clause in (111a) seems to be obligatorily extraposed to the end of the higher sentence in order to obtain the well-formed sentence in (111b).,Another example where extraposition applies obligatorily can be seen in (112).,

(That the father-in-law has come is good.) b*. Zvaita kuti vatezvara vauya,

(It is a good thing that the father-in-law has come.)

t / / / , / / / /(11.2)a, Vasekuru vati Pita akaenda kumusha.(Grandfather said that Peter went home.)

b. Zvanzi/ navasekuru Pxta akaenda kumusha*(It was- said by grandfather that Peter went home.)

c. Pxta. anzi^ navasekuru akaenda kumusha.(Peter was reported by grandfather to have: gone home.)

221

The underlined string im each sentence is the complement clause. In (112b) amd (112c) this complement clause has been extraposed* In (112b), first, the passive applies, interchanging the complement clause p A a akaenda. kumusha with the subject HP vasekuru, them extraposition applies obliga­torily taking the embedded clause to the extreme right end of the whole sentence. Im (112c), first, the passive applies: as im (112b); second, object raising replaces the promoun head of the subject HP with. Pxta; them extraposition applies obligatorily. Judging from: the examples above it is clear that extraposition is mot altogether optional. But subject postposing on the other hand is wholly optional. Besides, the string that is extraposed ceases to behave like a noun phrase in its new posi­tion. Hone of our tests, either for noun phrases in subject position or for noun, phrases in object position, will confirm, say, the extraposed complement clauses in (112b) and (112c) as noun phrases.

We might suggest the extraposition rule in (113) as a first approximatioEE. of the transformation that relates the sentences in (109) above*

(113)X -

HP

+H+CL8+pro

X

HP1 2 3 k

1 2 0 %+3

The tree in (110a) which represents the structure that underlies (109a) is properly amalysable in regard to the structural description of the extraposition rule and the resulting tree in (1.10b) is: of the form that is specified by the structural change of this rule*. The trees in (112a) and (112b) are repeated here as (ll^a) and (ll^b) respectively in order to illustrate how this rule operates,..

222

(114)a.

X -

COMP

VPHP

itamwana ipa

4

Exfir apo sition

X -

1

VP

COMP

VPNP

NP

ipa _ kuti __ 1+ sub

ita+N+cl8+pro

mwana

02

The formulation! of the rule of extraposition as given above in (113) is adequate to account for the relation between the sentences in (109)*But now let us see what the effect of this rule, as- presently formulated^ would be on a string like the following:

223

(113) ICuti, kuti shiamwari yake i<ti/ze/, z.vakatamibudz.a Bango zvakafadza /mapurisa,

(That that his friend escaped worried Bango pleased' the. police.} The deep structure for (113) is: something like (ll6a). Applying the extraposition, rule to (116a) will result in (11.6b.) *

(ll6)a*.

COMP

VPNP

COMP

NP VP

kuti </+sub2(/ .. yake

sharawari tiza tambu™ Bango fadza mapurisa dsa j

k1. 32

Extrapo sition

224

COMPHP

COMP VP

NP VP

NP

tamhu- Bango fadza mapu- kuti dza risa ^+su:

shamwari tiza yake- a

+N+CL8+pro

1 2 0 4 3

After applying later rules the resulting sentence is (117') which is un­grammatical.

i f f t t f / /(11?’) *Kuti zvakatambudza Bango zvakafadza mapurisa kuti shamiwari/ / /. ( /yake rtiz.ev

(*That it worried Bango pleased the police that his friend should escape.)

The umgrammatieality of (117) indicates clearly that our formulation? •" of the: extraposition rule in (113) was incorrect. Now let us modify a little the application of this rule to the structure in (116a)* Instead of. moving term: 3 to the extreme right end of as was done in (ll6h) , let us move it only to the extreme right end of The tree configura- tion which results is (118).

225

(118) 1VP

V FP

COMP

FP VP

COMP

FP VP

tarnbu- Bango kuti dza

kuti __ __ shamwari tiza fadza/+subj_«/ yake

+F+CL8+pro

mapurisa

After applying other rules the details of which are not crucial here, the sentence, we get is' (119) which is well-formed*.

(119) Kuti zvakatambudza Bango kuti shamwari^yake itize zvakafadza /mapurisa.

(That it worried Bango that his friend escaped, pleased.: the police*)

From what we observe in (118) and (119) it is obvious that when extraposition applies correctly, transporting a complement clause in subject position, it is sensitive to the constituents over which the complement clause moves. Stated in more precise terms, it can be said that it is sensitive to the structural distance that the complement clause may move. If we. compare the structural difference between the: grammatical sentence in (119) which is derived from (118) and the un­grammatical sentence in (117) which is derived from (ll6h) we are pro­vided with the information necessary for a proper formulation of the

226

rule of extraposition. In formulating this rule there is need to state that the complement clause, which is; moved by extraposition, may move only to the end of the. clause in which it originates. The rule of extra­position is then reformulated properly in the following way:

Rule (11) Extraposition ___

X -

S1 ' 2 3 h 5 = = >1 2 0 h+3 5

If we compare this rule with the one in (113) above we will see thatthey differ only in the presence of the fifth term on the right hand,which serves, to place a right boundary on the term; that is moved, namely,term 3* Th-e extraposition rule as reformulated derives the grammaticalsentence., in (119),, tut not the ill-formed sentence in (11?)..

If however Ross’s rather persuasive argument, namely, that allrules that move constituents to the right cannot move elements fartherthan the end of the clause which contains those elements, turns out tobe universal, the extraposition rule as formulated in (113) will be

13adequate. There will be no need to include the refinement above as it will be true of all right-ward movement transformations. This particular feature of right-ward movement transformations will need to be stated but only once in the general theory of language and will apply equally to all right-ward movement transformations.

1+F+cl8 - ^COMP s7 X - X_+prq_

, FP IP

13.. J.,R.Ross: (1967) Constraints on Variables in Syntax,

22?5-12 Pro Deletion. (non-cyclic)

TlfThe pronoun head, mantely +CL8

+proof mourn phrases with comple­

ment clauses is mover manifest in the surface structure, but is always deleted. Nevertheless the case for its existence in deep structure was argued in Chapter’ One. In the following sentences the pronoun head, though present in deep structure, does not show. Its trace can be seen in the concord zvi which is underlined.

/ / / / y / /(120)a. Kuti imrnhu angamhare pamwedzi zvinonetsa kunzwisisa.-(That a person can land on the moon is difficult to compre­hend.,)

/ / / / / / / /b. Zvinonetsa kumzwisisa kuti munhu angamhare. pamwedzi.(It is difficult to comprehend that a person can land on the moon.)

(121)a, Kama mwana achingochema-chema zvinoreva kuti haasi kunzwa« a k 4 & a .(If a child keeps on crying it means that it is not feeling well.)

b.. Zvinoreva kuti haasi kunzwa zvakamaka kana mwana a chin go -ch ema- eh ema.(It means that it is not feeling well if a child keeps on crying.)

Notice that in (120b) and (121b) extraposition has also applied. We may formulate the rule which deletes the pronoun head of complement clauses as follows:

Rule (12) Pro Deletion

X -

1

1

UP

+NI+CL8/15•fpro

2

0

NIP

2

228

This rule, which applies obligatorily, is ordered after extraposition and gender copying. Term 3 i*£ the structural description is given as either S or 0, If extraposition, which we said was optional in themajority of cases, has not applied, term 3 will be S, and this will bethe case with (l2oa) and (121a); but if extraposition has applied then term1 3 will be 0, and this is the case with (12ob) and (121b), That this rule must apply after gender copying is obvious enough. If this ordering is not followed the predicate of the containing sentence will be left without a subject concord and this will render the sentence ungrammatical#

3.13 Object Anaphora (non-cyclic)If in a sentence there are two identical noun phrases which

are not dominated by the same simple S-node aitd it is the case thatnot both of them are in subject position, one of the two noun phrases is reduced so that only a pro-form remains. This may occur either with­in one complex sentence as in (122) and (123) or across sentence boun­daries as In (12 -) below. The noun phrases which are identical as wellas their pro-forms are underlined,

✓ / / / / / / / / / /(122)a. *Vana^vanoziva kuti ticha anoda kurova vana^.(*The children^ know that the teacher wants to beat

the children^,)/ / / / / / / / / / /b* Vana vanoziva kuti ticha anoda kuvarova,

(The children know that the teacher wants to beat them,)(123)a, *Kana chibage chacho^ chakanak/ vanhu vanotenga chibage

chacho^,(*If the corn^ is good people will buy the corn^,)

/ / / / / / / / /b, Kana chibage chacho chakanaka vanhu vanochitenga.the corn is good people will buy it,)

(12£|-)a,i, Mwoyo adya chingwa chiye,(Mwoyo ate that bread,)

229

ii. Set Mwoyo adya chingwa c.hiye?(Why did Mwoyo eat that bread?)

/ / b.i. Mwoyo adya chingwa chiye.ii. Sei Mwoyo achidya?

(Why did Mwoyo eat it?)In these sentences it is the second of the two identical noun phrases which is reduced to a pro-form;. Note also that this operation is obli­gatory in (122) and (123), that is, when it occurs in one complex: sen­tence, but optional in (124), presumably because it occurs across a

T Jsentence boundary. This operation can be put into a formal rule as follows:

Euler (A) X - - X - NP2 - X1 2 3 4 5

1 2 3 “*4 “+pro 5

condition: 2 - 4 but these two noun phrases should not be dominatedby the samie S,

14. The concept of sentence boundary here appears to apply at the levels of (a) separate sentences as belonging to two different speakers as in (124) above, and (b) complex sentences which contain such conjun­ctives as: asif, nokuti, as illustrated in the sentences below.

(a)i. Sara, anoda Mxdzi asi Sekax achivenga Mxdzx chaxzvo,(Sara loves Midzi but Sekai hates Midzi very much*)

t / / / / , / / / , /ii. Sara anoda Midzi asx Sekax achxmuvenga chaxzvo.(Sara loves Midzi but Sekai hates him; very much.)

/ / / / / / / / /Chimbumiu anovhima tsoko nokuti anodya tsoko.(Chimbumu hunts monkeys because he eats monkeys.)

f t // / / / / / ■Chimbumu anovhima tsoko nokuti anodzidya. (Chdmbuniiu hunts monkeys because he eats them.)

230

Mow consider the following sentences:// / / / / / / /(123)a. Kama ndaiwana ndichapa mari yako kuna Taka.

When I get it I shall give your money to Taka.)b. Kama ndawana mari' yako'' ndichaipa kuna Taka,

(When I have got your money I shall give it to Taka.)(126)a. Zvawarisvlbisa cbigeza bhasikoro racho.

(Since you have soiled it, you wash the bicycle- now. )/ * / t , J . /b. Zvawasvibisa bhasikoro racho chirigeza.

($Ince you have soiled the bicycle, you wash it now.)In (125a) and (126a) it is the first of the two identical noun phrases which is reduced to a procformr, but in (125b) and (126b) it is the sec­ond noun phrase which is reduced. In (125a) and (126a) rule (A) above applies in reverse. The operation in (125a) and (126a) may be represen­ted in the form' of a rule thus:

(B) X - - x - np2 - X1 2 3 4 5

1 2 3 4 5+proIn (125a) and (126a) rule (B) has

sentence, and one is tempted to conclude that rule (B) operates only when an object noun phrase in an embedded sentence is reduced. But thefollowing sentences reveal that such a conclusion is fallacious.

i / f / / / / / /(127) Vanhu vanochitenga kana chibage chacho chakanaka.(People will buy it. if the corn is good,)

/ f / / / / /(128) Chirigeza zvawasvibisa bhasikoro racho.(Yon wash it now since you have soiled the bicycle.)

sinceMote that^tkese embedded sentences are functionally adverbials they appear in a position after the main verb in deep structure by convention.. In the two sentences immediately above rule (B) has applied to the noun phrase in the matrix clause. It would appear therefore that in Shona

231

the constraint found in English, namely, that if the first of the two identical noun phrases is reduced, it must he dominated by an embedded clause which does not dominate the second noun phrase, does not apply.

We can reformulate this rule in a slightly different way in anticipation of conflating it with rule (A).

MP! - X - np2 - Xk 3 2 1

3 2 1

3

3>

It can he seen that this rule is a mirror image of rule (A) above* Thatis, it applies in the opposite direction, from right to left. Our gram-mar is simplified if these two rules, which are identical save that theyapply in opposite directions, can be conflated. This is possible if weinvoke one of the provisions of Langacker's mirror image convention,which says that two rules can be collapsed if they contain exactly the

15same terms in their structural descriptions and output sequences, ^ I shall adopt his notation of marking the structural description of a mirror image rule with an asterisk. Thus, rule (13) collapses, rules (A) and (B) above.

Rule (13) Object Anaphora

*

*X - HP1 - X - RP2 - X1 2 3 ^ 3

1 2 3 “ | 3lipr2J

condition: 2 = if but the two noun phrases should not be dominatedby the same simple S,

What this rule says is that, when this structure index is met, either

13* Bor mirror image rules see R.W*.Largacker in "Mirror Image Rules 1: Syntax" in Language, vol.h3, Number 3 (1969) p.373 - 398.

232

term! 2 or term. i|., "but not both, is reduced to a pro-form'* As was mention­ed elsewhere above this rule is on the whole obligatory. It is optional only across sentence boundaries, Note that, however, across sentenceboundaries it applies only fromt left to right but no# in the other dir-

, . 16 ectxon,

3*1 Object Focussing (non-cyclic)When it is intended to focus attention on the object noun

phrase this is achieved by moving the object noun phrase out from its position, and placing it at the end of the sentence, then, substitutinga pro-forms in its place* This is illustrated below.

/ / / . / / / / / /(129)a. Kuba kwakatayxsa vakomama ava pfungwa chaizvo.(Stealing really made these boys lose their mind,)

b. *Kuba kwakatayisa^ pfungwa vakomana ava chaizvo.(^Stealing caused the mind to lose these boys very much.)

/ / / / / / / , / / /c., Kuba kwakavatayisa pfungwa chaxzvo vakomana ava.(Stealing really made these boys lose their mind.)

/ / * / , / A f(130) a.. Ifdaratidza mwana uye vatete vake.(I showed that child his aunt.)

L , * f / f *b* *Mdaratxdza vatete vake mwana uye.(*I showed his aunt that child.)

c. Ndamura t i d za vatete vake mwana uye.(I showed that child his aunt.)

16. In place of Rule (13) the following schema may be adopted with thesame results: X -- NP1 - X - *P2 - X

SD: 1 2 3 k 3

SC: (a) 1 2 3 P It +pro 5

(b.) 1 r z n 3L Jk 3jj-pro 1

For an argument against mirror image rules see Jorge Hankamer in "On the Hone existence of Mirror Image Rules in Syntax" in Syntax and Semantics, vol. 1, John P.Kimball (editir)

233

/ / /In (129a) the object noun phrases, vakomana ava and pfungwa, have a17fixed linear order. That is, they must follow each other in that se­

quence, the latter constituting a part or a fraction of the former.

The sequence of the two object noun phrases in (130a) , though not exhi­biting the inalienable relationship, is fixed also, that is, mwana uye. preceding vatete vake^ If tfo.es order of these object noun phrases is re­versed the resulting strings are ungrammatical as (129b) and (lJJOb.) show. But in (129c),, and also in (130c), because there are pro-forms, namely, va and mu respectively, which have been left behind, the first object!:NT in each sequence can: be moved comfortably to the end of the sentence yielding a grammatical sentence. The object NT so moved, or focussed, appears to be set apart from the rest of its clause. Besides being trans­ported from its position this object NP seems to be marked by a potential, slight pause or "comma intonation". This comma intonation sets off this transported noun phrase from, the rest of the clause.

The transformation which gives rise to object focussing appliesto structures which have the phrase marker in (131) and derives fromthem structures of the form in (132),

(131) e (132) S

X pro-form Y

The rule which is needed to derive (129c) from (l29a), and (130c) from (130a), is roughly as follows:

17# Buch object complements are treated at some length by N.C.Dem.betembe in Verbal Constructions in Korekore, an unpublished M.Phil,- Dissertation, London, (1969) p.33 - 64.

234

Rule (A) £c - NP - Y7S S

1. 2 3- = $

# ■ 2S

where MP is am object c.omplemient.This rule applies to a structure that is exhaustively dominated by S, adjoins r ■ a copy of the. noun phrase in. question within that structure

"[ Oas a sister to S,. and marks the original R'P as a pro-formi.Object focussing does not always consist of taking out a noun

phrase and moving it to the extreme right end of its clause as in the sentences in (129c--) and (130c). Object focussing can also optionally transport:, an object noun phrase to the left end of its clause* In the examples which follow the sentences in (129c) and (130c) above appear with the focussed noun phrase transported to the left end as in (133) and (134) respectively,

/ / > / / / / /(133) Vakomana ava, kuba kwafcatayisa pfungwa, chaizvo*(As for these boys stealing has really made them, lose their mind*,)

, V / / / / / / /(134) Mwana uye, ndamnratidza vatete vake.(As for that child I have shown him. his aunt.)

In (133) and (134) the comma intonation is even more pronounced thanf / t /previously, suggesting strongly that vakomiana ava and mwana uye are set

off from the rest of their respective clauses. In order to characterize the process that has taken place in (133) and (134) it would appear that there is need of a transformational rule of the following form:

P Ij+proj

18. For more information on dislocation transformations see Langacker

Ibid, p,589.

235

(B) n - NP - x7S “s£ 2 1

2 3 1+pro

=>

Mote that rule (B) has been formulated with a view to collapsing it with rule (A) later, Motice also that the meaning of (129c) and (133) is the same, so is that of (130c) and (134). Evidently rules (A) and (B) are similar in a significant manner. There is need, it seems, of establish­ing some notational device which will allow these two rules to be con­flated so that their similarity can be captured. Owing to the bidirect­ional character of object focussing the mirror image convention suggests itself. Notice that these two rules satisfy one of the provisions of this convention (see p.;'231 • In this case the output sequences are takento consist of four terms, and not three. The fourth term is term 2 which

e>u,fas pointed^earlier is set. off from: the rest of the clause. Allowing for this extension these two rules can be collapsed into one rule.

Rule (14) Object FocussingZ* - MP - y7

“si 2 3

i 2™ 3+proS

Tv* 2

S

This rule is to be interpreted as follows: it takes out of S the second term and places it by Chomsky adjunction either to the right end or to the left, end of S, leaving a pro-form behind.

3.15 Object Copying (non-eyclic)The object copying rule is a simple rule which applies only

after the rules of object anaphora and object focussing. What it does

236is to copy the class feature / + C L a n d the pro feature /+pro/ of an object noun phrase on to the. verb.., It is an obligatory" rule. The r&le may be represented formally as follows:

Rule (13) Object Copying

X

1

1

H yp 1V - I+CLi -

j+proj2 3

The application of thiLs rule' is very v ■- similar to that of the second subpart of the gender copying rule discussed in. 3*7 above. While, the latter copies the class feature on to the auxiliary, the former copies it on to the verb*

5..16- Noun. Prefix (segment) rule (non-cyclic)This rule, which is obligatory, applies only after gender copy­

ing and object copying. What it says is that whenever there is a term with a class feature Z+CLi^/, realize /jCLi^/ as a noun prefix before.-' the term from which it is copied. If we have, for instance, a noun withthese features ^punu, +CL7, spoon__/, this rule says that the-feature /+CL7:J must be realized as a prefix thus:

M * . « Mpunu+N ("pre fix!---- \ [jcL7 J + PUB.U

+N+CL7 +CL?*-*•

SpOOB

/ 1 M "*■»•

spoon.

When formalized this rule is:Rule., (16) Noun Prefix (segment)

X

1

1

Y+CL-

[""prefixl +2 *cli

H U H *

23?

After applying later rules:, including the one which spells £+CL?_7 as the prefix chi, the result is the noum chipunu (a spoon)* This ruleplaces the concords in their proper positions, that is, to the immediateleft of the terras from which they will have been copied.

3*17 Subject Postposing (non-cyclic)It is not. uncommon in Shona to find the subject noun phrase

transposed from its underlying position, which is before the auxiliary,to the end of the sentence. This transformation,it would appear, has the effect of focussing attention on the predicate. The (b) sentences in the following pairs have their subjects placed at the extreme right end..

(133) a. Hope dzabata amai vaye zvino ,(Sleep has overwhelmed that woman now.)

b*. Dzabata amai vaye zvmo hope.(Sleep has now overwhelmed that woman.)

(136)a* Chi^a inomhanya chaizvo.(lit. A cheetah runs indeed s= A cheetah is a fast animal.)/ / / /b. Inomhanya chaizvo chita*(It is a fast animal, a cheetah.)

t / / ✓ / / / / / /(137)a. Mudzimai uyu haana mwana anorohwa nomudzidzisi,(This woman, has no child who can be beaten by a teacher.)

_ __ t & / / a / / . y, / . , . *.b, Haana miwana anorohwa nomudzidzisi mudzimai uyu,(She has no child who can be beaten by a teacher, this

woman.)We can formalize this rule in this manner:

Rule (1?) Subject PostposingX - /RR - T 7 - Z

S ~S1 2 3 b

= 31 0 3+2 if

238

5.1.8' Rule Ordering and the motion of the CycleTwo types of rtitle ordering may be distinguished, namely, in­

trinsic and extrinsic ordering of rules.. While the former is generally accepted without much difficulty, the latter is a very controversial topic,. For some discussion on this matter, see Chomsky: 1965a; Postal: 1968b; G.Lakoff: 1968a; Ringen: 1972’.; and Koutsoudas: 1972. I shall not pursue the topic of extrinsic rule ordering at any length as it is not really crucial to ray main study. For the purpose in hand I have however assumed partialtextrinsic. rule ordering. What requires comment though is the notion of the cycle. I labelled some of the rules discussed in this chapter cyclic and others non-cyclic implying preference for a cyclic grammar. May I point out that this preference is to a large extent arbitrary for there is as yet no compelling evidence to make onechoose the cyclic grammar to, say, a linear grammar on empirical gxsnmds

19grounds. All that can be shown however with the evidence at hand is20that a cyclic grammar is as viable as a linear grammar. Another

reason, though less cogent, for preferring a cyclic grammar is that most of the study in the generative transformational model has been carried out within its framework.

In a cyclic grammar the rules (i.e. those that are cyclic) apply first to the most deeply embedded S, then to the next higher S, and so on until the highest S in the configuration is processed. The primary motivation for the cyclic principle of rule application requires basically that some rule R^ apply both before and after some other rule R .. Schematically, in terms of a derivation, this is illustrated in (138).

19* Cyclic and linear grammars are discussed by John P.Kimball in Syntax and Semantics , Vol.I, John P.Kimball (ed.), p.63 - 80,20. See John Grinder in. Tl0n the Cycle in Syntax" in Syntax and

Semantics for the motivation of the notion of the cycle in a grammar. Refer also to foot-note 6 p.138.

239

(138) Underlying Structure

Intermediate Structure

Intermediate Structure,Intermediate Structure

1

32

by rule R by rule by rule R^

Surface Structure by rule RThis derivation involves two applications of R^ and a single applicationof R^* There is no intermediate string in the derivation which meetstlie structural description r ' for R . except the string created by the Tfist application of R^, After this first application of there is no

intermediate string which is properly analysable with respect to R^ except the output string of R^. What we have then are two applications of R± which are made possible only if R^ applies between them. The trans­formational rules which apply in a cycle are partially ordered extrinsi- cally among themselves. For illustration I shall consider the interaction of (a) the passive and object raising rules; (b) the reflexive and subject raising rules; and (c) gender copying and subject raising rules, 5*18,1 Passive and Object Raising rules

Consider the sentence in (139) and also compare it with (?2a),/ ./ // /T , ✓ / /(139) Bhora rxnodiwa naticha kuti ritamibwe nevana,

(a ball - is wanted - by the teacher - that - it is played -by children := The teacher wants the football to be played by the children,)

The structure which underlies (139) is provided in (lifO).

2^0

PreS HP VP

passive ticha V HP

dl H S

VPPreS IP AUXkuti

passive vana V HP

tamba bhora

After applying the passive to S^ we get the structure in (Ikl)*

CUhl)

PreS HP

passive ticha V

da H

+H+CL8+pro

COMP

VPAUXkuti HP

bhora

t amib.wa lie van a

Then object raising applies to give us the tree in (1^2)•

2 ZO-

(142)

PreS IP

passive ticha V WP

bhora COMP

VPkuti AUX

taimbwa ire vana

Fote that bhora. is b o w a direct object ini S^# By applying the passive the (14-3)

in. the configuration we get is (1^3)•

bhora V

diwa COMP

kuti AUX

maticha

tambwa mevama

After extraposition of S the result is then (l£|if).

242

(144

COMPbhora

diwa naticha kuti AUX

V PP

tambwa mevana

In deriving (139) from (140) first the passive was applied* Its output created the structural description for the object raising rule, the out­put. of which in. turn produced a structure, which was properly analysable for a second application, of the passive rule. The interaction, of these

f ■ "v*i y k ttwo rules describes how the noun phrase bhora which is at the hwfrb end of the sentence in (72a) has moved up the tree to be the subject of the topmost S in (139) •• The interaction of these two rules is quite consist­ent. with the notion of the cycle as described above..

dispenses with the notion of cycle we can still derive (139) from (14o). Transformational rules are ordered as in a cyclic grammar. According to this approach a transformational rule is applied first to the most deeply embedded S, then to the next higher S which meets its structural descri­ption until the highest S is reached, A rule is constrained to apply with an upward orientation in a given configuration. Then, the next rule is applied in the same fashion, until all the rules relevant to a given structure have been applied.

in (140). I shall assume that the order of application of these rules is passive and then object raising. The first, application, of the passive (i.e. to S^) results in the same tree structure as in (141) above. The

Nevertheless:, by adopting the linear grammar approach which

Mow let. us see how this works with respect to the structure

243

second application of the passive yields the structure in (145)* (145)

+M+CL8*pro

GOMP

VPkuti FP AUX

bhora V

diwa

PP

tambwa mevana

naticha

Allowing for its reformulation (the details of which, are not essential here), the rule of object raising is considered next* Its application to the tree above will result in the structure in (146).

(146)

bhora

VPkuti AUX

diwa naticha

tambwa mevana

Extraposition applies, next.

244

(14-7)

VPHP 8

bhora PP COMP

diwa naticha kuti AUX

tamhwa nevana

The configuration in (14-7) is the same as that in. (14-4-) from which we said (139) derived.. Notice however that the notion of an upward orient­ation offi a structure is independent of the cyclic - non-cyclic principLe,

5*18*2 Reflexive and Subject Raising rulesThe interaction of the rules of reflexivization and subject

raising helps to motivate the notion of the cycle. To this end considerthe sentence in (14-8)*

f t t f f f f / / ((148) Zuze zvaazvicheka pamupimbiri azvikonzera kunetswa nenhunzi.-(Zuze, because he cut his shin, has caused himself to be bothered by flies.)

Hereunder is the rough underlying structure of this sentence.(14-9)

NP.1Zuze kunetswaCOMP

VPAUX

Zuze

cheka Zuze pamupimbiri

Zh5

The structure in S i s properly analysable for reflexivization. We. also/ /notice that the verb -konzera belongs to a class of verbs which allow

subject raising to apply on their abstract subject UP* Application of this rule to yields the result (150).

(1 0)

NP

COMP

VPAUXzva

konzera Zuze kunetswanenhunzi

I Zuze^ I+refie.

+ cheka pamupimbiri

Note that after the application of subject raising there are now two occurrences of Zuze in which paves the way for a second application of the reflexive rule. Thus the two applications of the reflexive and the one of subject raising which comes in between them strongly support the notion- of the cycle as described above.

Adopting the linear grammar and assuming that the order of application of these two rules is reflexive and then subject raising, it will be seen that reflaxivizatiom will apply to S o n l y , but not to

In its structural description will not be met. So this ordering of these two rules will not help. Possibly in a linear grammar the order­ing of these two rules is subject raising and then reflexiv&zation. Ap­plying subject raising first, we obtain the structure in (l^l).

2if6

(151)

NP S

COMP

VPAUX

konzera Zuze kunetswanenhunzi’

Zuze pamupimbiricheka

We them apply the, reflexive rule. The problem we encounter immediately is that S2 is. no longer properly analysable for the reflexive rule al­though S.,, now is. This reordering of these two rules does not help us

1 j*either to d^Lve (lif8) from (lZf9)* From this it may be deduced that a linear grammar is weaker than a cyclic grammar*

Nevertheless, if raising (subject or object) is regarded as a copying rule which copies the subject MP of the embedded, clause- into the next higher clause, the argument above which is in favour of the cyclic principle, is easily countered. Copying rules are common in a generative transformational grammar (cf. gender copying and obj'.ect copying)• Allowing raising to be a copying rule and assuming the order­ing of these rules to be raising and then reflexive, (1^8) would then be derived in the following manner:

(a) apply raising to UP2 by copying that term, from S2 into

ZWI

(152)

WPJL

Zuze COMP

zva RP2 AUX

Zuze

konzera Zuze kunetswa nenhunzi

cheka Zuze pamupimbiri

(b.) we then apply the reflexive rule to S„, marking NP, for reduct-2 5ion. to the reflexive form,

(153)

MP.l

COMP konzera ZuzeZuze

I Zuze- [ + cheka pamupimbiri I -t-raflexl

kunetswanenhunzi

(c) next apply the reflexive rule again, this time marking NiP for reduction to the reflexive: form,.

24-8

(1540

s

COMP S2

AUX VP

V HPZuze

kunetswanenhunzi

EZuze + cheka pamupimbiri •eflexj

The only other rule of concern to us here is Equi-MP deletion which deletes; MP^ from S as in (155)*

(155)

s

COMP S

VPAUXE Zuze [ + konzera kunetswa +reflexI nenhunzi

EZuze. 1+ cheka :e;flex|

pamupimbiri

In this way sentence (14-8) is correctly generated. Note that in this derivation there is one application of the rule of raising followed by two applications- of the reflexive rule. The latter are contiguous. The. application of these rules in this derivation is therefore mom-cyclic.*

249

5.18.3 Gender Copying and Subject Raising rulesThe interaction of these two rules yields similar results. To

this end consider and compare the sentences in (156) whose underlying structure appears in (157).

(156)a. Nyay/ zvayadai^zvichanetsa.b. Nyaya zvayada^ichanetsa.

(Since the affair has taken this turn there will be a difficulty.)

(157)

+N+CL8+pro

COMP

FP AUX VPzva

netsa

nyaya dai

Let. us consider the cyclic principle first. Gender copying is applied first to S^. This will eventually result in the concord in zvayadail Next, since the structural description for subject raising is met, we choose to apply it and the configuration which results is (158).

(158) S.

COMPnyaya netsa

250

The head noun of the subject NP of the higher S is no longer

flyaya which has among its features /+CL9^_7* Now gender copying isapplied, but im the higher S this time.. This will eventually result in

/ / / -the concord i in ichanetsa. In this way (156b) is generated. If subjectraising had not been applied, but only gender copying, the final result would be (156a). Note that in order to derive (156b) from: (157) it is essential that the application of these two rules follow this order: gender copying, subject raising and gender copying. Applying, for insta­nce subject raising first in the series will leave without a subject NP, and consequently the' rule of gender copying- will block,

Now let us consider the linear grammar principle. Let us alsofirst

assume^that these two rules are ordered this way: gender copying, sub­ject raising.. If. gender copying is applied to both and S^ first and then subject raising after, we finally come up with (159)•

(159) Nyaya^ zvayadai^ zvichametsa.This sentence by itself is grammatical, but is ungrammatical if consid­ered as having been derived from (157) by two successive applications of gender copying followed by one of subject raising. The drawback is that after applying these rules in this order we end up with a situation in which the raised subject and its new predicate are not in concordial agreement. The. grammatieality of (159) is a result, not of subject rais­ing, but of complementizer placement.

Let us now consider the other ordering of these two rules.: subject raising, gender copying. Applying subject raising first immedi­ately raises a problem since it leaves without a subject, and conse­quently gender copying is effectively blocked in S H o w e v e r if subject raising is considered as a copying rule as was the case in 5*18.2 above, (156b) can be derived from. (157) without difficulty.. The order of appli­cation of these rules will be: subject raising,, gender copying. Subject raising will apply followed by two applications of the gender copying rule..

251

What has been shown in this subsection is that there are at least two different ways of looking at how transformational rules apply to configurations which contain embedded sentences* The privilege which the cyclic principle has enjoyed over the years is after all not unassai­lable. Judging from, the foregoing discussion in this subsection, neither the cyclic principle nor the linear principle is preferred over the other on empirical grounds.

5*18.^ SummaryApart from establishing that the cyclic principle is no less

powerful than the linear principle, the account in 5*18*1 - 3 has shown that some rules are ordered, at least partially, among themselves. Of the rules discussed in this chapter the following have been shown to be cyclic: passive, reflexive, gender copying, object raising and subject raising*. Others which appear to be cyclic are: pseudo-passive, agent deletion and Equi-NP deletion. Since the pseudo-passive operates on the same structural description as the simple passive, its cyclic applica­tion is obvious enough not to require demonstration. Agent deletion is a rule whose application depends on a previous application of either the passive or the pseudo-passive rule, and since the latter are cyclic, it would appear that it too is cyclic.

The passive, the pseudo-passive and the reflexive rules are 21mutually exclusive, while agent deletion will apply only after either

of the first, two rules he ve applied. Gender copying is ordered after the passive and the pseudo-passive. Objiect raising, subject raising and Equi-MP deletion can only apply after the gender copying rule, but among themselves they are mutually exclusive. Below is a tentative ordering of these rules within a given cycle.

21* For the occurrence of the passive and pseudo-passive formatives with one and the same verb see foot-note 7 P* 18?

$2

Gender'

Agent Deletion

'Passive■Pseudo-passive

Reflexive

S>Obj>ct RaisingSubject Raising

Nsfequi-FP Deletion The linear ordering within the cycle follows the direction of the arrows. None of the other rules, it would appear, is cyclic,. For none of the non-cyclic rules is ordered before any of those that are cyclic. Some of the non-cyclic rules are nonetheless ordered among themselves. For instance, extraposition applies; before pro deletion; object copying must apply after either object anaphora or object focussing; however object copying as well as gender copying applies before the noun prefix rule.

6.0 THE GRAMMATICAL FUNCTION OF COMPLEMENT CLAUSESIn Chapter One it was shown, that sentences may be embedded. in

other sentences*, These embedded sentences or complement structures are introduced by complementizing morphemes which are listed in subsection 2.10* The occurrence of these complementizing morphemes is specified by the phrase structure rule in (32) in Chapter Four, It was also shown that these complement structures are of two types, namely, those that are dominated by NP (see 1.1.0) and those that are dominated directly by VP (see 1*2*0), In this chapter I would like to show the grammatical function of these complement structures: to the extent that this is pos­sible, I shall consider first noun phrase complementation and then verb phrase complementation.

6*1 Noun Phrase ComplementsNoun phrase complements, as we saw, are structures with pro­

perties which are characterized by the phrase structure rule NP ^ N(S)

which was given, in (49d) in Chapter One., These sentential complements may function grammatically as subject and as direct object of verbs.Just as noun phrases may be objects of prepositions in, say, Latin or English, in Shona noun phrases may also be objects of elements like, /ik/and /na/, I shall now illustrate each of these grammatical roles whichcomplement clauses may assume,

6*1,1 Complement structures in subject positionA complement clause in a subject position is represented by

the following structure:

234Occurring in this position, a complement clause may be introduced by any of the six complementizing morphemes which are listed in 2*10, Here­under are examples of complement clauses in subject position* The comple­ment clauses in question are underlined.

A* Complement clauses introduced by kuti /jr indicative// / / / / / / / /Cl) Kuti umorara zvako kusvika masikati pamba pavatezvara zvino-

nyadza.(That you sleep until midday at your father-in-law's home isshame ful.)

/ / i / / t / / /(2) Kuti Podzorimwa achasvika mangwana zvinondifadza.(That Podzorimwa will arrive tomorrow pleases me.)

, , / / / / / / / / / / /(3) Zvakaita hazvo kuti Murungwenl akabva asvikawo panguva iyoyo.(lit. It happened by itself that Murungweni arrived at thattime = Luckily enough Murungweni arrived at that time.)

Note that in (3) the subject complement clause is exfcraposed to the endof the sentence.

Complement clauses may also occur as subject In sentences inwhich the subject concord either does not appear at all or is controlled

1by a verb complement NP*/ / / / t f t(4-) Kuti Mwedzi akasvika nezuro kunyepa.

(That Mwedzi arrived yesterday is false.)/ / / / / / / / l /(3) Kuti akarovha hasa nokuti airwara hachisi chokwadi.

(That he did not. come to work because he was ill is not the truth.,)

B. Complement clauses introduced by kuti /+ subjunctive/(6) Kuti ndidzokere kudongo kwandakambobva zvapata muhana mangu.,

(For me to go back to my former village site won't be accept­able to me.,)

/ / / / / / j t i f f /(7) Kuti tipote tichinooma vabereki kumusha zvakanaka chaizvo.

1. For more information in this connection see p.34 “ 36.

255(For us to go home regularly and see our parents is a good thing.)

/ / / / ' f t / / /(8) Kuti am buya vanwe tii isina mukaka zvava ku r amb a•(For grandmother to drink black tea is now difficult.)

, s , / / / / / / / / f /(9) Kuti Tongai auye mangwana zvinongondifadza.(For Tongai to come tomorrow will still please me.)

The comment made above in (A) with regard to the control of subject con­cords equally applies to complement clauses introduced by kuti /+ subjjV*

/„ . / / / /(10) Kuti titize umbwende.(For us to run away would be cowardice.)

(11) Kuti Roda asyike pamusha pababa wake iye akambotiza mukumbo/ / haisi nyore.

(For Eoda to go to her father's home after she had eloped is not an easy matter.)

C.. Complement clauses introduced by kana /+ participial/or kuti /+ participial/

, . / / / / / / / / / / ✓ /(12) Kana mombe dzako dzarara mudoro make zvichamusvota chaizvo.(If your cows were in his garden the whole, night it is going

to worry him very much.)/ / / / / / / / /(13) Kana Tembeni apunza ndiro imwe chete hazvina mhosva,

(If Tembeni has broken only one plate it does not matter.,)(14) Kana Gordo akarovha hasa kwesvondo rese zvichamuomera,

(If Gordo absconded from work the whole week things vd.ll be difficult for himi.)

(15) Kuti wakatengera amai rokwe zvakanaka*f

(If you bought your mother a dress it is fine.)D., Complement clauses introduced by kana /+ consecutive/

, . v / / / / / / /(16) Kana ukaoia chiva chichifamba zvinoshura.(If you see a puff-adder in motion it implies ill luck.)

(1?) Kana ukabatwa muhapwa zvinonyeredza.(If some one touches your arm-pit it tickles.)

256

(18) Kana mwana akatuka amai vake zvinoera.(It is shocking for a child to scold his/her mother.)

i f f f / / /(19) Kana ukagara. uchidya nyama chete zvinofinha.(If you eat meat only every day it will get monotonous.)

(20) ^ana ukapota uchibata-bata tubasa zviloioyamura.(If you undertake small pieces of work it will help (yam})

The complementizing morpheme kana in the sentences (16) - (20) may be deleted optionally without effecting any semantic change.

(21) Ukaona chiva chichifamba' zvxnoshura. (see 16)(22) Ukabatwa muhapwa* zvinonyeredza. (see 17)(25) Mwana akatuka amai zvinoera. (see 18)(2 -) Ukagara uchxdya nyama chete zvxnofinha, (see 19)(25) Ukapota uchxbata-bata tubasa zvinoyamura. (see 20)

(E) Complement clauses introduced by ku /+ imfinitive*7/ / /(26) Kufadza munhu hazvisi nyore.

(To please a person is not easy.)(27) Kuona tsuro xchxombera zvxnoomesa mate mukanwa.

(To see a rabbit clapping is breath-taking, (lit. it dries the saliva in the mouth.))

, nx t f t t i * / / /(28) Kupopotedzana nevadiki zvxnofumura.(To quarrel with youngsters is degrading.)

(29) Kunamata bedz.i hakuna kukwana.(Praying alone is not sufficient.)

It is not. infrequent for the subject of the complement clause in this case to be deleted. One other observation worth pointing out here is that the sentences in which complement clauses introduced by ku /+ imfiru/ are subjects tend to express generalizations, and in consequence the complement clauses have indefinite subjects, hence these are usually deleted. In the sentences (26) - (29) the subjects of the complement clauses are deleted. But in (30) and (31) below the subjects are present*

257

(50) Mwana kutuka amai zvakaipeu(For a child to scold its mother is deplorable.)

(31) Mbudzx* kudya mufe'njie hufana nyina.(For a goat, to eat the cabbage-tree shows resemblance to the mother.)

The last example is a proverb. Notice that when an infinitive is turned into a predicate ku sometimes becomes hu.

(F) Complement clauses introduced by z.va /+ participial/(32) Zvawatora mari^ yake zvichakukonzera kutaura.

(Taking his money as you have done will get you into trouble with him (lit. will cause you to speak).)

(33) Feso zvaava kuchovha makasa zvi^chamurunzira kuba.(This gambling for money which Feso is. now practising will tempt him to steal.)

(3k) Matai zvaapera zvi^chandipinza muchikwereti.(The fact that my tyres are worn out will lead me into debt.)

(35) Zvaauya nhasi kudai zvinore^a kuti basa akape^dza.(Coming today as he has; done indicates that he finished the work.)

6,1..2 Complement clauses as direct object.In this position, in a sentence complement clauses, like non-

clausal noun phrases, may occur by themselves or alongside indirect objects. This is illustrated in the tree structures in (a) and (b) res­pectively below*

(a) * (b)

V NP NP

258

Indirect objbcts Shona are marked by ku/kuna and, to a lesser extent, by pa/pana. In deep structure these indirect objects are considered in this study to occupy a position after the direct object* Note that the occurrence of indirect objects, but not that of direct objects, is opt­ional in such sentences. This is exemplified below where the indirect objects are underlined,.

(36)a* Kudef akapira bhora kuna Yoyo,(Kuda passed the ball to Yoyo.)

b. Kuda akapira bhora,(Kuda passed the ball,)

c. *Kuda akapira kuna Ycfyo,(*Kuda passed, on to Yoyo.)

/ / / / /(37)a. Ndakwereta marl iyi kumukoma wangu.(I borrowed this money from my brother.)

b. Ndakwereta mari^ iyi.(I borrowed this money,)

/ / / /c. *Ndakwereta kumukoma wangu.(*I borrowed from my brother,)

Generally, the tests for indirect objects are the same as; for direct objects which were provided in. subsection 1.2,2.

More often than, not, however, by an optional transformational rule which I shall refer to as the Indirect Object Movement rule, the indirect object sheds the formative ku/kuna or pa/pana and moves to a position immediately before the direct object as shown in (38) and (39)*.

(38) Kuda^ akapira Yoyo bhora,(ICuda passed Yoyo the ball.)

(39) Ndakwereta mukorna wangu marl iyi.(i borrowed from my brother this money.)

In sentences: in which the direct object is a complement clause this in­direct object movement rule applies almost always in Shoma.

259

Embedded* sentences serving as direct objects may be introduced by kuti /+ indicative/, kuti /+ subjunctive/, kana /+ participial/ or ku /-]- infinitive/.6.1*-2.1 The sentences in this section contain direct objects, but

not indirect, objects.A*. Complement clauses introduced by kuti ,/+ indicative/

(Zf 0 ) Jioko akafembera kuti achahwina mari kumahachi.(Joko guessed well that he would win some money on the horse racea.)

(Jfl) Siyaso akabvumia kuti akatadza.(Siyaso admitted that he made a mistake.)

/ / / / / / / / / / / /(A2) Taka anodaira kuti zvinhu zvese zvichunaka kumusha.(Taka believes that everything will be all right at home.)

(A3) Peni^ aramba kuti akazvitaura.(Peni denied that he said it.)

B. Complement clauses introduced by kuti /+ subjunctive/(iflf) Rwizi^ anzvenga kuti aende kumombe.

(Rwizi has gojs. out of looking after the cattle.)/ / / / t / f * f(A3) Churu akaedza kuti abope mombe akabva atungwa.

(Churu tried to inspan the oxen and. was in consequence gored.) (A6) Vanatsa kuti vaende mvura isati yanaya.

(They did well to go before it rained.)(A7) Mutongwa akatsika-tsika kuti ataure chokwadd.

(The accused was hesitating as to whether or not to tell the. truth..)

(AS) Dare rakasunga kuti varoyi vatandirwc mhiri kwaMasikandoro (The court ruled that the witch-doctors should be expelled across the Masikandoro river.)

C. Complement clauses introduced by kana £+ participial// / / / / / / / /(A9) Tete vari kuda kuziva kana vana vauya.

260

(Aunt wants to know whether the children have come.)(50) Gogo abvunza kana muhari mu chine doro.

(Grandmother has asked whether there is still some beer in the po t.)

(51) Handichayeuka kana takambosangana kare.(I don't remember whether we have met before.)

/ ✓ / / / / , / / / '(52) Marimo ari kuedza kufunga kana zvataurwa namudzvitx zvingaitika.(Marino is trying to figure out whether what the district commissioner said can be. done.)

D. Complement clauses introduced by ku /+ infinitive/(53) Mukomana ari^kudziidza kutyaira motokari.

(The boy is learning to drive a car.)(5b) Chirau ajaira kunwa hwahwa^ hwepachena.

(Chirau is used to drinking free beer,)(33) Kobi. akanonoka kunotora vana kukxrechn.

(Koni was late in going to fetch the children from the creche.)(36) Zvakapinda murairidzi muchiko^ro vanef vakabva vanyarara kutaura.

(As soon as the teacher entered into the classroom the child­ren stopped talking.)

E, I have discovered only one example so far of a complement clausein object position which is introduced by kana /+ consecutive/,

i f / / / t(37) Tichaona haazive kana Pesi akasvikako nhasi.(Tichaona does not know whether Pesi will? xk get there today.)

Even then the status of the complement clause as an object is doubtful since all the tests for object noun phrases, except one, yield results the acceptability of which is questionable.

(a) pseudo-cleft test;(38) ?Kana Pesi^akasvikako nhasi ndizvo zvaasingazive Tichaona.

(Whether Pesi will get there today is what Tichaona does not know.)

261

(b) passive test;:(59) ?Kana Pesi^ akasvikako nhasi hazvizivikanwe maTichaona..

(Whether Pesi will get there today is mot known by Tichaona*) (e) object anaphora test;

( 60) ?Tichaoma haazyizive. ( i * e kana- ge.si^ akasvikako nhasi)(Tichaona does not know it. (i.e. whether Pesi will get there

today.))But the interrogative pro-form test produces a positive result.

(61) Tichaona haazive chii? (reply: kana Pesi^akasvikako nhasi) (Tichaona does; not know what? (reply: whether Pesi will get

there to day ))The fact that the complement clause in (57) is; unclear, coupled with the paucity of examples, tends to suggest that complement clauses which are introduced by kana /+ consecutive/7 do not feature as- direct objects.*

No example was attested of a complement clause introduced by zva / + participial/ in direct object position.

6.-1.-2.-2 Examples of sentences with both direct and indirect objectsA* Complement clauses introduced by kuti /+ indicative/

i i i l i t /(62) Mdaudza Sara kuti Taka akauya.(I told Sara that Taka had come.)

(63) Ticha akanyevera Matenzeni kuti akapasa.

(The teacher hinted to Matenzeni that she had passed.)

(6h) Posani abvunza Charwe kuti mukoma wake amodzoka rinhi.(Posani asked. Charwe when her elder sister would return.)

(65) Baba vakaraira sekuru kuti amai vari kurwara.(Father sent word through uncle that mother was ill.)

B.. Complement clauses introduced by kuti /+ subjunctive/

(66) VaChiutsi^ vaida* kumanikxdza miwana^ kuti aite kosi yaasingade. (Mr.,Chiutsi wanted to force his child to take a course whichshe did not want.,)

262

(67) Amai vakatsindidza mwanasikana wavo kuti asataurire munhu* t tzvechikomba chavo.

(The mother entreated her daughter not to tell anyone about her (mother’s) boy-friend*)

(68) Mubatwa akaraira murnwe mukomana, kuti vabereki vake vauye:p / / /ku zomut sikinura.

(The accused sent word to his parents through a certain young man to the effect that they should come to bail him out.)

C,. Complement clauses introduced, by kana /+ participial/7These have been difficult to come by..

(69) Toni^ abvunza Sekesai/kana Chipo aclzoka^.(Toni enquired from Sekesai whether Chipo was back.)

/ / / / / / t / t(7 0 ) Mdichakuyeuchidza kana npuva yekudya yasvika.

(I shall remind, you when it is meal time.)D. Complement clauses introduced by ku /+ infinitive/

( 7 D

(72)

(73)

(7<+)

(75)

X » 3

the type marked by the element /$t/. (For this type of phrase see also

Appendix A.) This phrase occurs a immediately after the verb in (76),

Tr i’ , / /, / y / , m , / / / / i / f / / . /VaGondo vakaraanrkidza Toko kuroorwa nemukomana waakanga asi-

ngade.

(Mr.Gondo forced Toko to get married to a boy she did not love.)/ ✓ / / / / / /Tete atuma Magi kunotora badza. kumunda.

(Aunt sent Maggie to go and fetch a hoe from, the field.)/ / / / / / f tMari akabatsira Nyika kutakura mbatya dzake.

(Mari assisted Nyika to transport his goods.)t / * t / /Eega kufurira mwana kuba.

(Do not lead the child into stealing.)Musikana uye akakonzera Zuze kurohwa nematsotsrl (That girl caused Zuze to be beaten up by hooligans.)

Complement Clauses as object of /ir/In Ghona almost every verb may be accompanied by a phrase of

263

immediately after the direct object in. (77) j and immediately after the second objject in (78)* The /ir/ phrases; are underlined in these sentences*

(76) lesi auyira m a n yake*(Jesi has come for- her money*)

/ / / f(77) Makoni atyorera rura yaSani nharo *(Makoni broke Sani’s ruler out of sheer cantankerousness.)

(78) Sara anyimira vana chekudya misikanzwa yavo*(Sara refused to give the children any food because of their

mischief,)

The rough deep structures of these three sentences are given in (79) ~(81) respectively.

(79)

Jesi

ir mari yake

(80)

Makoni V

tyora rura yaSani

ir nharo

nyima chekudya vara P NP

ir misikarzwa yavo

In all these three structures the Applied transformational rule will theninsert obligatorily /ir/ before the terminal vowel of the verb to give

/ / / / / / / / ✓ . .us the verbs -uyira, -tyorera and -nyimira.. The morpheme /ir/ is distri­buted in accordance with the rules of vowel harmony in Bantu languages*.

That is, /ir/ occurs if the penultimate vowel of the verb stem, simrple or extended, is /i, a, u./, and /er/ occurs if the penultimate vowel is

/e, o/.These purposive phrases, as I propose to call the phrases

which are marked by /ir/,, may express purpose or the reason for which the action (or state) of the verb is accomplished (or entered into) as

in (76) - (78)* This type of phrase may also signify the person or thing' for which the action (or state) of the verb is done (or entered into)

as in (8 2) - (8 4 ) *t t * i(82) Ndatengera Pumbidzai bhora iri,

(I bought this ball for Rumibidzai.)(8 3) Kugara paro zvakarakira iwe, kwete inif*

(For us to stay here is all right for you, not for me..)(84) Warpisira shamwari^yako nyaya^ yake*

(You have spoiled your friend's case for him.)When the /ir/ phrase has this connotation it is necessarily moved fromc its position to a position immediately after the verb as in (8 2) - (8 4)* In view of this difference in connotation, perhaps the two occurrences

265

of /ir/ should not be equated.

The tests for a noun phrase which is object of /ir/ are the

same as for direct objects which were outlined in subsection 1.2.2.

Taking (85) as an illustration, it can be shown that kuroya is a noun

phrase object of /ir/.

(8 5 ) Taruvinga akarambira mukadzi kuroya.(Taruvinga divorced, his wife because of her witchcraft.)

(a) pseudo-cleft;

Kuroya ndifco k waakar amib ir a mukadzi^ Taruvinga.

(It was witchcraft: which made Taruvinga divorce his wife.)

(b) passive;

Kuroya kwakaramhirwa mukadzi^ nai'aruvinga.

(witchcraft - was divorced for - a woman - by Taruvinga =

T a r u v i n g a 1s wife was divorced by him because of witchcraft.)

(c) interrogative pro-form;/ / / / / / (Taruvinga akarambira mukadzi chii? (reply: k u r o y a )

(Why did Taruvinga divorce his wife? (reply: because of witchcraft))

(d ) o bj e ct anapho r a ;

Taruvinga akaaviram.bi.ra mukadz/. (i.e. k u r o y a .)

(Taruvinga divorced his wife because of that. (i.e. witchcraft))

Since one cannot perform an action on behalf of or to the

advantage or detriment of an object noun phrase of the type (N: + S ) ,

n o u n phrase complement:: clauses will not be found in /ir/ phrases with

the latter connotation. It is therefore with the former connotative ex­

pression of the /ir/ phrases that we shall be concerned here. The noun

phrase complement clauses which have the privilege of occurring in these

/ir/ phrases are mainly those introduced by kuti /+ subjunctive/ and

ku /+ infinitive/,

A* Complement clauses introduced by kuti /+ subjunctive/* / / / / / a / / ' / / /(0 6 ) Vamwe vakaenda kumunda, asi Toko akasrira kuti achengete ambuya

266

y / /vairwara.

(The rest of the family went to the field, but Toko remained

behind in order to look after grandmother who was ill.)

(8?) Baba vakarimira pizi kuti vatengese.(Father grew the peas in order to sell.)

/oox „ / / / / / / , /, ( . / . / / , . / / /(8 8 ) Vamwe varume vanoroorera vakadzi vazhinji kuti vawane vanova-/ / / / /batsira basa chete.

(Some men marry many wives just in order to get people to

help them, do their work.)

(89) Joko aida kukandira VaMhesvu chibhakera kuti vapore kuvhaira/kwa v a i i t a .

(Joko wanted to give Mr.Mhesvu a blow in order to stop him.

boasting.)

The /ir/ phrase occurs immediately after the verb in (86); after the

direct object in (87) and (88); and after the second object in (89).B. Complement clauses introduced by ku /+ infinitive/

(90) Vana vamirira kunzi chiendai^ n a ticha.

(The children are waiting for a word from the teacher to the

effect that they may now go.)

(91) H e m naFombo vano f a n r a kudya nyama y e h u k u .

(Reni and Fombo feel happy when they are having chicken.)

(92) Mvura iri kupenyera kuzorara ichinay^.(There is so much lightning that it will rain during the night.)

(93) Murairidzi akatandira mwana kurovha chikoro.

(The teacher expelled the child because o f his/her irregular

attendance,)

(9h) Posani akaratidzira amhuya musikana wake kusvot/sa Rozi^

(Posani introduced his girl-friend to his grandmother in order

to annoy Rose.)

As above, the /ir/ phrase occurs immediately after the verb in (90) -

267

(9 2);after the direct object 1b (93); sond after the second object in (94)* C. A few examples have been attested also in which the complement

clause is introduced by kuti /+ indicative/./ / / / / / / / / /(95) Feso ndakamurovera kuti anotuka vakuru.

(I beat up Feso because of being rude to his elders.)(96) Zuze ndakamuzvondera kuti haadi* kutengawo hwahwa kana tiri

kubh^.wa »(I detest Zuze for his reluctance to buy beer when, we are at

the beer hall.)I have not found examples; of /ir/ phrases' ill which the comple­

ment clauses are introduced by either kana /+ participial/, or kana 1+ consecutive/, or zva /+ participial/.

The /ir/ in these phrases is: not infrequently deleted option­ally. In the examples which follow /ir/ is deleted..

(97) Varnwe vakaenda kumunda asi^Toko akasara kuti achengete ambuyjSivairwara. (see 86)

(98) Baba vakarima pizi kuti vatengese^. (see 8?)/ / / / / / / / t * / / /(99) Vamwe varume van.oroora vakadzi vazhinji kuti vawane varnova-

batsira basa chete., (see 88)/ / / / i t t / t /(100) Joko aida kukanda VaMhesvu chibhakera kuti vapore kuvhaira

kwavaiita. (see 89)(101) Posani akaratidza ambuya musikana wake kusvotesa Rozil(see 94)

But applied to (85)> (90) and (93) this deletion transformation produces negative results as shown by the ungrammaticality of (102) - (1 04).

(102) ^Taruvinga akaramba mukadzi^kuroya.(1 0 3) *Vana vakamira^ kunzi chiendai^ naticha., . / / / / / /(104) *Murairidzi akatanda mwana kurovha chikoro

In some cases the sentences in which /ir/ appears are less acceptable than those from which it is deleted.. The (a.) sentences in

the pairs below are less acceptable, than the (b) sentences.

268

' / / f t / S ,(105)a, Murwere akabudira panz.e kuti afurwe nemhepo.b., Murwere:. akabuda panz.e^ kuti afurwe nemhepo.

(The patient went out of the house in order to get some

fresh air.)/ / / t / / / /(106)a. Sand ari kudyira nyama kuti akore.t f t i j / /h. Sani ari kudya nyama kuti akore.

(Sami is eating meat in order to gain, weight.)f f / / / / / / * J(10?) a.. Mutyairi akacheukira kuti aone kana miapurisa ait ever a iye»./ / / / / / / / / . / /b. Mutyairi akacheuka kuti aone kana m apunsa aitevera xye.

(The driver looked back in order to see whether the police

were trailing him.)Further investigation is needed here in order to ascertain more preciselywhen this deletion, transformation applies and when it does not.

That /ir/ has been deleted in (97) - (101) and also in (105b), (106b) and (10?b) receives further confirmation from the pseudo-cleft

test. I shall illustrate this point using the sentences in. (97)> (99)and (101).. The (b) sentences are ungrammatical because /ir/ does not

appear.(1 08)a. Vamwe. vakaenda kumunda asi Toko, kuti achengete amhuya,

ndizvo s v a d a s a W - .(lit*. The rest of the family went to the ftLeld but Toko in

order that she might look after the grandmother is why she remained behind = The rest of the family went to the field but Toko remained behind in d r ; order to lookafter the grandmother.)

/ / / / / / / tb. *Vamwe vakaenda kumunda asi Toko, kuti achengete arnbaya,/ / /ndizvo zvaakasara*.

(109)a* Vamwe varumie, kuti vawane vanovabatsira hasa chete, ndizvo

zwavanoroorera vakadzi vazhmja.(lit. Somerr c men in order that they may get people to

269

help them in doing their work that is the reason why they marry many wives = Some men marry many wives in order that

they may get people to help them do their work.)/ / / / / / / / / / /h*. *Vamwe varurne, kuti vawane vanovabatsira basa chete, ndizvo

/ t / f t / 'zvavamoroora vakadzi vazhinji./ / / / / / / / / /(110)a. P o s a n i , knsvotesa R o z i , ndiko kwaakaratidzxra amhuya/ / t tmusikana wake.

(lit. Posani, i n order that he might annoy Rose, that was

the reason why he showed his grandmother his girl-friend

= Posani showed his grandmother his girl-friend in order

to annoy Rose.)/ / / / / / / / / / /b.. *Posani, kusvotesa R o z i , ndiko kwaakaratidza amhuya musi-

kana^ wake.In the sentences immediately above the concord zvo in ndxzvo is clearly

a Reflex of the noun prefix of class 8 , indicating that these clauses

which are underlined have a noun o f class 8 as their head noun. That

there are /ir/ phrases in (9 7 )> (9 9 ) a^d (1 0 1 ) is shown by the presencef / y* / / fof /ir/ in zvaakasarira in: (1 0 8 a), /er/ in zvavanoroorera i n (1 0 9 a) and

... / / / ,/xr/ in kwaakaratidzira. in (110a). The pseudo-cleft transformation is

very reliable for discovering prepositional phrases with deleted prepo­

sitions in Shona..

Another test which will confirm the strings underlined i n (97)

- (1 0 1 ) as well as those in (1 0 9 b), (1 0 6 b) and ( 1 0 7 b) as prepositional

phrases with /ir/ deleted is the interrogative pro-form transformation.

In order to elicit as. answers the strings which are underlined, /ir/

must be present in the questions being asked. I shall take (98), (100),

(109b), (106b) and (107b) for illustration. As above the (b) questions

are ungrammatical because /Ir/ is absent.t t t f / / i f / /(111)a. Baba vakarimira chii pizi? (reply: kuti vatengese)

(Why did father grow the peas? (reply: in order to sell)), / / / / / / b. *Baba vakarima pizi chii?

270

* • / / / / t / ,(1X2)a.. Joko aida kukamdira VaMhesvu chibhakera chii? (reply: kuti

vapore kuvhaira kwavaiita)(Why did Joko want to give Mr*Mhesvu a blow? (reply: in

order to stop him boasting))t / / / i t fb, *Joko aida kukanda VaMhesvu chibhakera chii?

/ N f i t / / / / / , / / / /(113)a* Murwere akabudira panze chii? (reply: kuti afurwe nemhepo)(Why did the patient go out of the house? (reply: in. order

to get, some fresh air))f f , *, *b, ^Murwere akabuda panze c h n ?

i f f / / i t / /(llif)a, Sami ari kudyira nyama chii? (reply: kuti akore)(Why is Sami eating meat? (reply: so that he may put on

weight))t t t i t /b, *Sami ari kudya nyama chii?

i / t t t t/ t t(113)a. Mutyairi akacheukira chii? (reply: kuti aone kana mapurisa/ /aitevera iye)(Why did the driver look back? (reply: in order to seewhether the police were trailing him))

/ / / ✓ /b, ^Mutyairi akacheuka chii?

Notice that these purposive objects may also occur after other

complement structures,

(116) Rwizi^ amzvemga kuenda kumunda kuti ago^wera ari kutaundishipj.. (Rwizi avoided going to the field so that he might then spend the day at the township,)

That the strings kuenda kumunda and kuti ago^swera ari kutaundishipj are direct object and purposive object respectively is confirmed by the tests for object noun phrases,

(a) pseudo-cleft;/ .,-N / t / t / /(115?) a, Rwizi kuenda kumunda ndiko kwaamzvenga kuti agoswera ari

kutaumdishipi,

(To go to the field is what Rwizi avoided so that he might

271

them spend the day at the township,)/ / / / / / / /b.,. Kuti agoswera ari kutaumdishipi ndizvo zvaamzvemgera kuenda

/kumunda Rwizi,

(That he might spend the* day at the township is the reason

why Rwizi avoided going to the field,)

(h) pseudo-cleft and passive combined;z' / z t / /(118) a, Kuenda kumunda ndiko kwanizvengwa maRwizi kuti agoswera ari

/kutaumdishipi,

(Going to the field is what has been avoided by Rwizi so

that he might then spend the day at the township,)/ / / / / / /b, Kuti agoswera ari kutaundishipj ndizvo zvanzvengerwa kuenda

/kumunda maRwizi,

(lit,. That he might spend the day at the township it is it

what was, avoided for to go to the field by Rwizi =

That he might spend the day at the township is the reason

why Rwizi avoided going to the field,)

(c) interrogative pro-form;

(11.9)a,. Rwizi amzvemga chii*kuti agoswera ari kutaumdishipi? (reply

kuenda k u m u n d a )

(Rwizi avoided what in order that he might then spend the

day at the township? (reply: going to the field))

b, Rwizi^amzvemgerei^kuenda kumunda? (reply: kuti ag/swera ari

kutaumdishipi)

(Why did Rwizi avoid going to the field? (reply: in order

that he might then spend the day at the township))

Further examples of sentences like the one i n (116) appear below,N s / / / ✓ / ✓ ✓ /(1 2 0 ) Akaregera kunwa hwahwa kuti achemgetedze rnari,

(He stopped beer drinking in order to save money,)

(1 2 1 ) Ndakahatsira Mashiri kudyara mbeu dzake kuti iye ago'zondiba-/ /

tsirawo,

272

(I helped; Mashiri to plant, his seeds in order that he might help me. in t u m k)

(122) Magi^ aka.ro vha ku.enda kuchikoro kuti agonoo'na bhaisikopo..(Maggie, absented herself from, school in order to go to thedLnema. )

6/1 .if Complement Clauses as Object of /raa/ (i.e.. instrumentalphrases)

Like purposive objects, instrumental objects occur in the form of prepositional, phrases.. Whereas the preposition was /ir/ in pur­

posive phrases, with, instrumental phrases it is /na/. For noun phrases which are objects of /na/ only two of the tests for object noun phrases as outlined in 1.2.2 are applicable. These two test© are the pseudo-cleft and the interrogative pro-form. For illustration consider the sentences

in (123)., l i t i f t /(123) Tembeni aneta nekuti auya. adzokera.

(Tembeni is tired, with this coming and going.)

(a)' pseudo-cleft;t i t / i f f /Kuti auya. adzokera ndizvo zvaaneta nazvo Tembeni.

(This coming and going is what. Tembeni is tired with.,)(b) interrogative pro-form;

Tembeni aneta nechii? (reply: nekuti auya adzokera)(Tembeni is tired with what? (reply: with this coming and going))

Note that in the case of' the pseudo-cleft, transformation, as the object noun phrase of /na/ is moved to the beginning of the sentence, it leaves behind a pro-form. In the case at hand the pro-form is zvo.,

Only complement structures introduced by kuti £+ indicative/*

kuti /+ subjunctive/" and ku /+ infinitive/ have been attested in instru­

mental phrases. These phrases indicate the instrument by which the act­

ion (or state) of the verb in the containing sentence, is accomplished

273

(or entered into)* In the sentences below the instrumental phrases are

underlined.A. Instrumental clauses Introduced by kuti /+ Indicative/

(12i+) Sekuru vachafara nekuti wauya..(Uncle will feel happy that you have come.)

(125) Monika ari'kuyaura nekuti akatukwa nado.(Monica is upset at Jo's having scolded her.)

(126) Chipo aizvirumbidza^ nekuti anogona muchikoro asi akazokundi-/kana*.

(Chipo used to indulge in. self praise saying that she was bright in class but she finally failed.)

B* Instrumental clauses introduced by kuti /+ subjunctive/(127) Ndine urombo nekuti usvik/ pano nhasi pasina chekubata*.

(I regret it very much that you have come today when I have

nothing to eat.)

(128) Ambuya vanofara nekuti ndiende^ ndiroo'na' muuun*una kujeri kwaari.(Grandmother is happy for me to go and visit my young cousin

in prison where he is*.)C. Instrumental clauses introduced by ku /+ infinitive/

(129) Mazuva atatamira muGota mombe, dzakapera nekufal (When we moved into Gota area many cattle died.)

(130) Jameti aneta mekukuya.a. (Jeneti is tired because of grinding.) orb. (Jeneti is fed up with grinding.)

(1 3 1) Vana^ vakarasa mombe nekutamba.(The children lost track of the cattle through playing.)

(132) Kumyange zvairambidzwa Dehwe akapa vapanduki zvokudya n/kutya. (Although it was unlawful Dehwe gave the outlaws somie food

through fear.)

In some cases' /na/ can be left out readily without prejudice to the meaning of the sentences as in the sentences below,-

(133) Tembeni aneta kuti auya adzokera. (see 123)f t t f(13k) Sekuru vachafara kuti wauya. (see 12k)

(135) Kdine urombcf kuti usvike pano nhasi pasina chekubata. (see 3L2(136) Mazuva atatamira muGota mombe dzakapera kufa. (see 129)(137) a. Rond/ ak^.para mhaka nekuroVa mupurisa.

b. Roni^ akapara mhaka" kurova mupurisa.(Roni committed a crime by assaulting a policeman.)

. n. / / / / / / / / / / / / /(138)a. Murume akakahadzika nekuona mbatya dzake dzabiwa.b. Murume akakahadaika kuona mbatya dzake dzabiwa.

(The man was struck dumb to find his property stolen.)

That the strings kuti auya adzokera in (133)j kuti wauya in

(13k)> kuti usvike/* pano nhasi pasina chekubata in (1 3 3)? kufa in (136),. kuruva mupurisa in (137b) and kuona mbatya dzake dzabiwa in (138b) are. /na/ phrases from which /na/ has been deleted is further confirmed by the pseudo-cleft and the, interrogative pro-form tests, tor illustration

I shall consider (1.33) - (133) only.(a) pseudo-cleft;

t t * / / / / t(139) Kuti auya adzokera ndizvo zvaaneta nazvo Tembeni.(Coming and going is what Tembeni is tired with.)

(lhO) Kuti wauya ndizvo zvavachafara nazvo sekuru„(That you have come is what uncle will be happy about.)

(lkD Kuti usvike pano nhasi pasina chekubata ndizvo zvandine uro-J *mbo nazvo,(That you have come here today when I have nothing to eat is what I regret very much.)

(h) interrogative pro-form;

(lk2)a. Tembeni aneta nechii? (reply; kuti auya adzokera)

(Tembeni is- tired with what? (reply; of coming and going))

2 75

t t / Jb.. *Tembeni aneta c M i ?(1^3)a-* Sekuru vachafara mechii? (reply: kutl wauya)

(What will please uncle? (reply: that you have come)) b, * Sekuru vachafara chii?

(lZfZj.)a* Une uromibcf nechii? (reply: kuti usvik/ yamo nhasi pasina/ / /chekubataj

(What are you sorry for? (reply: that you have come heretoday when I have nothing to eat))

/ / /b* *Une urombo chii?

In other cases this deletion transformation will be blocked* Applied to (131)j (132) and (lh7a) it results in the ungrammatical sen­tences in (lh3)j (lk6) and (lAf?b) respectively.

(1^5) *Vana vakarasa mombe kutamba.(1^6) *Kunyange zvairambidzwa Dehwe akapa vapanduki zvokudya kutya.

✓ ✓ / t / / / / /(lh7)a* Aoma shumba akadonhedza pfurrto nekudedera.(On seeing the lion he dropped the spear because of tremb­

ling. )ft / / / / t /b*. *Aona shumba akadonhedza pfurno ku.dedera.

In the examples which follow the (h) sentences in which /na/has been deleted are preferred to the (a) sentences in which it appears*

/ / / / / / /dko)a„ Ndakashama nekunzwa kuti rnusha wakatsva./ / / / / /b*. Ndakashama kunzwa kuti rnusha wakatsva..

(I was surprised, to hear that the village had got burnt.)

(12f9)a. Sekuru vanofara nekuti upote uchisvika pano. b* Sekuru vanofara kuti upote uchisvika pancf.

(Uncle will be happy if you come here often.)(130) a* Tine urombo nekuti wakashurwa zvakadaro..

b. Tine urombo^ kuti wakashurwa zvakadaro.

(We are sorry that you suffered such a mishap.)

Now consider the following pairs of sentences.

276

(1 5 1 ) a,. Jeneti aneta nekukuya.

(i» Jeneti is fed up with grinding*) or

(ii, Jeneti is tired because of grinding*)/ t ib. Jeneti aneta. kuk u y a ,

(Jeneti is fed up with grinding,)

(1 5 2 ) a* Mombe iyi yapera nekurumrwa nezvxkwekwe*.

(i. This cow is covered all over with ticks (but not

necessarily emaciated)) or

(ii. This cow is emaciated by ticks*)t f * Ab* Mombe xyi yapera kurumwa nezvxkwekwe.

(This cow is covered all over with ticks (but not necessa­

rily emaciated^)

Note that (151a) and (152a) are ambiguous as the Englisjptranslation

shows, but (151b) and (152b) are not*. Notice also that it is the first

reading of sentence (a.) in each pair which tallies with that of the

corresponding (b.) sentence*. To the extent that they are semantically

equivalent, the sentences in the pairs in (1 5 1 ) as. well as (1 5 2 ) are

transformationally related* This relationship may be described in terms

of the optional deletion rule mentioned above*.

If, on the other hand, it is the second reading of the (a)

sentences: that is considered, then the (a) and (b) sentences in (1 5 1 )

as well as those in (152) are not related transformationally* The ambi-/ /guity in (1 5 1 a) and (1 5 2 a) is attributable to the verbs -neta and -pera

which carry each, at least, two different meanings* Another interesting

pair is the one in (1 5 3 ) which, on the face of it, the. sentences

appear to be related structurally by the deletion transformation already

referred to above*.

(1 5 3 )a* Ndanzwa mekutemwa nemusoro*

(I have a severe headache*). m , / / / / / /b* Lidamzwa. kutemwa nemusoro*.

277

(I felt a headache, for- a while (but mot necessarily a severe

one).)

Apart, from; their difference in. meaning the; pseudo-cleft transformation also shows that these sentences are not transformationally related. This is illustrated in (15k) and (1 55) respectively.

(15k) Kutemwa nemusoro ndiko kwandanswa mako. (cp 153a)(It is a headache that I am suffering from very much.)

(155) * Kutemwa nemusoro ndiko kwandanzwa rako.. (cp 153h) but Kutemwa nemusoro ndiko kwandanzwa. (cp 153h)

(It was a bit of a headache that I had.)In (153t) the complement clause kutemwa n/mu s or o'' is a direct object of the verb -nzw I , and not a /na/ phrase. The difference in meaning between

the sentences; in (153)> once more, is attributed to the different mean­ings which the verb -nzwa^ carries. In other words this-, difference in meaning is attributed to the subcategorizational statement of the verb -nzwcu That is, in (153&) -nzwa requires a /na/ phrase complement, but in (153b) it requires a direct object noun, phrase.

Notice that it is possible to have two /na/ phrases, of the instrumental type in one and the same sentence.

(156) Mombe dza^era n/kufa nekurumwa n/mhesvi.(The cattle have dwindled very much in number through death

caused by the bite of the tsetse-fly.)

(1 5 7) Vanhu vanzwa nekurwara nekurumwa neumhutu.(Many people are suffering because of an illness caused bythe bite of mosquitoes.)/ / £ / / / / /(158) Temba apenga nekuti Ruzvidzo naPasi vakaba mari yake/ / / / / J 1 / / / / tnekuvaona vachinwa hwahwa hwendhari ivo vasingashande.

(Temba alleges, strongly that Huzvidzo and Pasi stole his money because he saw them drinking beer which they had bought, but

they are not working.)

278-

8,1*5 Complement clauses, introduced by kana /+ participial/ and

zva /+ participial/ may also serve as time adverbials of a sort. The complement clauses in question are underlined, in the examples which follow,

(1 5 9) Mungae^nde zve'nyu kuntba kana mtapedza basa ramuri knit a,(You may go home when you have finished the work you are doing.)

(160) Kana ndapedza G,C,E. yangu ndinoda^ kuz<j>ita ukotil (After I have gone through my G.C.E. I would like to do a

nursing course:,)(181) Wadzanai^ ndamuona zvaauya,

(I saw Wadzanai when she came*)(162) Tembo zvaabatwa vanhu vombofurwa nemhepo,

(Now that Tembo has been, arrested, people will breathe freely,) These strings which are underlined, seemi to be noun phrases functioning as adverbials. As time adverbials, complement clauses introduced by

either kana /+ participial/ or zva /+ participial7 tend to control agree­

ments of class 16 rather than those of either class 8 or class 15 as the pseudo-cleft transformation will show. The sentences in (163) - (165) correspond, to those in (1 5 9)» (161) and (162) respectively,

( 1 6 3 ) Kana mapedza basa ramuri kuita ndipo pamungaende zvenyukumb/,.(Only when you have finished the work you are doing may

you go home,)/ / / / / / / / / / /b,. -'Kana mapedza basa ramuri kuita ndizvo zvamungaende zvenyu

kumba,, . N t / // / , / /(16t)a. Wadzanai zvaauya ndipo pandamuona,

(When Wadzanai came is when I saw: her,)/ / / / / f fb„. * Wadzanai zvaauya ndizvo zvandamuoma,

(165)a*. Tembo zvaabatwa Slanhu. ndipo povombofurwa n/mhepo,.

(Now that Tembo has been arrested is the time people will breathe freely,.)

279

/ i i / / / /b* *Tembo avaabatwa vanhu ndizvo zvo vombofurwa nemhepo*/If the interrogative pro-form transformation with rinhi (when), which

/is used in place of chii with time adverbials, is applied, these under­lined strings will, be confirmed as noun phrases* As an illustration

consider (159) a m (1-61).' 1 f * * * ■ ,(166) Tingaende zvedu kumba rinhi? (reply: kana mapedza basa ramuri-.-

khita)(When are. we free, to go home? (reply: when you have finished

the work, you are' doing.)), x i / t t , / / .(1679 Wadzanai wamuona rinhi? (reply: zvaauya)

(When did you see Wadzanai? (reply: when she earns))

It seems clear therefore that the strings underlined in (159) - (162) are noun phrases.

In addition to functioning as subject, complement clauses in­

troduced. by kana £+ consecutive/ may also serve as concessive adverbials.(168) Kana mapnrisa akapatarika zvawo sei Kongo havambomuonal

(Ho matter how the police, search for Kongo they will neverfind- him*.)

/ / / / / / / /(169) Tichavaona chete kana vakapfuura nekuno*.(We shall certainly see them if they pass through here*.)

That the adverbials in (168) and (169) are noun phrases is sopported by the pseudo-cleft test*. As was the case with the time adverbials mention­ed. above the complement, clauses here control the concords of class 16rather than that of either class 8 or class 1 5 *

, , t / / / / / /(170) Kana mapurisa akapatarika zvawo sex Kongo ndipo pavasingambo-/ / /muona*

(Ho matter how much the police search for Kongo they will not get him*.)

(171) Kana vakapfuura nekuno ndrpo chete patichavaona.

(If they pass, through, here it is only then that we shall

280

see them*)

6.1*6 Complement Clauses in both subject and object positionsIn Chapter Five we saw that complement clauses could oacur in

b o t h subject and object positions in one and the same sentence. In this

section I should like to pursue this question further, and show the

interdependency of the complementizers. My investigation in this regard

seems to indicate that there is n e e d to group the verbs, which have the

privilege of occurring w i t h such multiple noun phrase complements, into

two, viz. those represented by z-r -reva (mtean; say), - taura (speak, talk),

and -ratidza (show) which I will label as verbs of "meaning", and thoset /

which are represented by -ita (do, cause), -konzera (cause, be result of)* * ,and~runzira (cause), which I shall call verbs of "causation". Since the

occurrence of complement clauses in positions other thait subject and

direct object is relatively little understood at the moment I shall con­

fine my study in this section to complement structures that may occur

in subject and direct object positions only.

6.1.6.1 Verba of "meaning"a

W i t h verbs of "meaning" the complement clause in^subject^.posi­

tion may be introduced by any of the complementizers save kuti /+ indie./5

but the object complement clause may be. introduced by kuti /+ indicative/

or ku /+ infinitive/, the latter being accompanied by possess!visation

(for this possessive see subection 2*,12). Below are illustrative sente­

nces with verbs of "meaning".

A. kuti £+ subjunctive/ iu subject position(i) kuti /+ subjunctive/ verb — ■— kuti /+ indicative/ — -—

(172) Temba kuti asvike pamushgf patezvara wake zvinoreva^ kuti/ / / / /akabvisa chimanda.

(For Temba to set foot at his father-in-law's, home implies/ /that, he has paid out the chimanda beast in the marriage

negotiations.)

281

(173) Kuti Masawi abvuim/ kubhadhara imiwe mari^ yaari^ kunzi akabat f / / /zvinoratidza kuti ndiye akaitora.

(For Masawi to agree to pay some of the money which he is alleged to have stolen shows that he is the one who took it.)

(17A-) Kuti asvike nguva ino asati' asvika zvinotaura kuti haachasvika/ ✓ / / /nokuti mabhazi ose apera.

(That he has not yet arrived until now means that he will not longer come since all the buses have gone..)

(ii) kuti { jr subjunctive/ v e r b ku /+ infinitive/-----t i i t / / / / / /(175) Kuti Pesi atakure tsaga rechibage zvimoratidza kugwinya kwake.

(For Pesi to lift a bag of mealie meal shows his strength.)B. kana /+ participial/ in subject position

(i) kana /+ participial/ v e r b kuti /+ indicative/-----(176) Kana usina kumuo'na zvinoreva kuti atlza.

(If you did not see him it means that he has run away.)

(177) Kana Juru akariga mn.sasa uye zvlnoratidza kuti murume a ne/simba.

(If Juru felled that musasa tree it shows that he is a strong man.)

(178) Kana wagogodza Rozi akarega kudaira zvinoreva kuti axie chiko- mba mumba^ imomo.(If you knocked at the door and Rose did not answer it means that she is with her boy-friend in that room.)

(ii) kana / + participial/ v e r b ku /+ infinitive/-----(179) Kana Juru akariga'' musasa/ uye zvinoratidza kuve kwake mu rum./

ane/ simbal (see 1 7 7)(180) Kana wagogodza Rozi^akarega kudaira zvinoreva kuve kwake

/ / /neehikomba mumba imomo» (see 178)In (176) - (180) kuti can be used readily in place of kana.

C. kana /+ consecutive/ in subject position

282

(i) kana /+ consecutive/ — v e r b kuti /+ indicative/7 -----(181) (Kana) akasauya nhasi zvakare zvinoreva kuti pane zvakasara

/ s /zvoxtika kumr sha.(If be does not come again today it will mean that there issomething that happened at home after my departure,)

i i f / / / X(182) (Kana) ambuya vakanorarako zvinotaura kuti Magorimbo arwra/

chaizvo.(If grandmother spends the night there it will imply that Magorimbo is seriously ill.)

fenna £+ consecutive/ v e r b ku /+ infinitive/ — — ---, « , , i f f / / f t / / /(18 3 ) (Kana) ukaenda kumba kwaMashirx zvakare xwe wakambotukwako

/ f f / / f t / /zteakadaro zvingaratidze kuve. rombe kwako.-(If you go to Mashiri’s house again after you have been scolded that much it would show your stupidity..)

(l8if) (Kana) vavakidzani acakapota vachiyamurana zvinopangidza/ / t /kunzwana kwavo.

(If neighbours help one another from time to time it is a sign that they are on good terms.)

In (181) - (l8h) kana is optional as was pointed out in subsection 2.6*^D. ku /+ infinitive/ in subject position(i) ku //* infinitive/ v e r b kuti /-i- indicative/ ------

/ O N * t t t f / f t f(183) Ramudzo kusvika nhasi zvinoreva kuti akapedza basa.(For Ramudzo to come today means that he finished his work.)

(186) Fdoro kusvipa kumieso^ kudaro zvinoratidza kuti kunwa'✓ /chikokiyana.

(For Ndoro to have such a dark face implies that he is tak­ing skokiaan beer.)

(ii) ku /+ infinitive/ v e r b ku /+ infinitive/------(187) Kuzadza kwake' dura kunoratidza kushanda kwake n/simba.

(His filling the barn with grain shows how hardworking he is.)

283

(188) Mukuwasha ku s ao mb e p an a naatezvara zvinoreva kusawirirana kwavo,

(For a son-in-law not to clap his hands when greeting his father-in-law indicates their not getting on well together.)

The infinitive serving as subject may be a free infinitive as in (18$) or an infinitive which is necessarily accompanied by a possessive as

in (187)..E. zva /+ participial/ in subject position(i) zva /+ participial/ verb ----- kuti /+ indicative/ -----

/ O N / ✓ / / / / / / / / /(189) Zvaasina kudaira kugogodza kwawaita zvinotaura kuti ane/ tchikomba murnbamo.

(Since she did not answer your knock it means that she is with her boy-friend in the room.)

(190) Tembeni zvaari kungorutsa-rutsa zvinoratidza kuti ava^/ . .nepamuvin.

(Since Tembeni is vomiting from time to time it shows that she is pregnant.)

(ii) zva /+ participial/ v e r b ku /7 infinitive/-----/ / // / i / / / / / / /(191) Zvaasina kudaira kugogodza kwawaita zvinotaura kuve kwake

nechikomba murnbamo. (see 189). / / ' / / / / / / /(192) Tembeni zvaari kungorutsa-rutsa zvinoratidza kuve nepamuvirikwake. (see 190}

6-.1.6.2 Verbs of ''causation"With verbs of^causation" the complement clause in subject

position may be introduced by any of the complementizers except kuti /+ indicative/, but the object complement clause may be introduced by

kuti /+ subjunctive;/ or ku /+ infinitive/ only. The infinitive may be free or accompanied by a possessive.

A. kuti /+ sub.junctive7 in subject position(i) kuti /+ subjunctive/ v e r b kuti / + subjunctive/-----

T_ . . / f , { s f / / / / /(193) Kuti mugare muchirwa zvmokonzera kuti murambane.

281

(For you to be always fighting will eventually lead youinto divorce.)

f t t f t / / / / S ' S(19A-) Kuti ndidaokere kuchikoro zvinokunda kuti ndiramba ndichiitabasa reumabharani.(For me to go back to school is better than continuing as

as a clerk.)

(195) Kuti uitewo rako gejo zvinobatsirar kuti urege kunetsana /nababa pakurima..

(For you to have your own plough will help you avoid trouble with your father at ploughing time.)

(ii) kuti /+ subjunctive/ v e r b ku /+ infinitive/------. . . / / / / / / / /(196) Kuti mugare muchirwa zvinokonzera kurambana. (see 193)

f t f * / / ✓ r / / /(197) Kuti ndidzokere kuchikoro zvinokunda kuramba ndichiita basa/ /r eumiabhar ani. (see 19^)

B. kana /+ participial/ in subject position

(i) kana / + participial/ v e r b — kuti /h- subjunctive/----(198) Kana Nyasha anga asingatere makore o'se aya zvichaita kuti

/ f t /anyimiwe munda.

(If Nyasha has not been paying his tax all these years this

will cause him. to be denied a plot of land.,)t / t t t / / t t(199) Kana mombe. dzaSoko dzadya mumunda maGavi zvichakonaera kuti

/ / / /vavengane.(If Soko’s cattle have laid waste Gavi’s crop this will make them enemies.)

(ii) kana /+ participial/ v e r b ku /+ infinitive7-----/ N / / / / / ✓ / / / / /(200) Kana Mago achipota achienda kunonwa hwahwa muniisha zvichamu-

/ / / / / x\runzira kudzingwa chikoro.

(If Mago is in the habit of going out to drink beer in the

villages this will cause him to be expelled from school.)

, * / ' * s s / / / / /(201) Kana Mushonga achiita basa rekuranda vanhu mari yavo zviclia-

mukoBrzera kugara ac hit ever wa nemapurisa.

(If Mushonga is fond of robbing people of their money this will result in his being closely watched by the police.)

C. kana /+ consecutive/ in subject position

(i) kana / + consecutive/ — -— verb ----- kuti /+ subjunctive/ ----. , N ✓ / / ✓ / / ✓ / / / V /(202) (Kana) ukagara uchirwa memukadzi zvinokonzera kuti murambane.,

(If you fight with your wife often this will bring aboutyour divorce.)

(203) (Kana) munhu ukava' muranda wekachasu zvinoita kuti chitaka/ / / / / / /chako chiore pakugumisira.

(If you become a slave of kachasu spirit this will cause you to have cancer of the liver in the end.)

(ii) kana /+ consecutive/ v e r b ku £+ infinitive/------/ V / v / / ✓ / / / / / / /(20k) (Kana) murume nemukadzi vakagara vachirwa zvinokonzera

kurambana.(If a man and his wife are always fighting this will leadto their divorce.)

, X , V / / / / / / / > ■ / / ✓ ✓(205) (Kana) ukajaira kuchovha makasa zvinorunzira kufa uchibaiwaneb a n g a .

(If you develop the habit of gambling for money this will lead y.,.:r to your death by stabbing.)

Notice that in (202) - (205) kana is optional.D. ku /+ infinitive/ in subject position(i) ku /+ infinitive/ v e r b kuti /+ subjunctive/-------

(206) Kugara chete kunoita kuti muviri urukutike.(Idleness makes the body weak.)

/ / / / / / / / / / f / /(207) Kushanda nesrmba kunobatsira kuti munhu ubudirire.(Working hard helps one to be successful,.)

286

. .. / / / i / i f(20o) Kuchovha makasa kunorunzira kuti miun.hu uite ubhinya. (Gambling for money will lead one to become a thug.)

(ii) ku /+ infinitive/ v e r b ku /+ infinitive/ —, % / / / / / f(209) Kunyarara kunokunda k u r w a .

(Keeping quiet is better than fighting*), v / / / / / /(210f Kusanzwa kunokonzera kutambudzlka pashure *

(Lack of understanding is the forerunner of troubles.)

E. zva /+ participial/ in subject position

(i) zva /+ p a r t i c i p i a l / v e r b kuti /-i- s u b j u n c t i v e / -----/ / / / / f / / / s f /

(211) Mukomia zvaakona kuenda kumusha zvichaita kuti ini ndiendeko*

(Since my elder brother has failed to go home this will make

mie go there instead.)f / ' ' / / ✓ / / / ✓ / / / /(212) Zvawaramba uchitamba nemuramu wako nemutowo iwoyo zvichaku-

konzera kuti utaudzane nemukadzi i z v i .

(Since you continue to play about with your sister-in-law in

that way, this will bring about misunderstanding between

you and your wife.)

(ii) zva /+ p a r t i c i p i a l / v e r b ku /+ infinitive/ 6 --— —, x / / / / ✓ / / / ✓ ✓ / / / f(213) Zvawaramba uchitamba nemuramu wako nemutowo iwoyo zvichaku-

/ / / ✓ / /komzera kutaudzana nemukadzi izvi. (see 212)

6,.. 1*6*3 Distributional Chart of ComplementizersThe distribution of complementizers in multiple complement

clauses is summarized below*. The sign /+/ in the table below indicates

a combination that is allowable.

A distributional Chart of Complementizers

in object position

f4o*H43■HTOOft-Po0)•r-a

Pi•H

6*2 Verb Phrase Com.plernentsThe case for verb phrase complements was argued in Chapter One..

In that chapter it was stated that the complement clauses which do notcome under the domination of NP, that is, those which do not pass thetests for noun phrases, are considered to be directly dominated by VP.In this section I should like to look at the different types of VP ofwhich these complements are constituents. I must point out at the out­set that the position with regard to verb phrase complements does not

appear to be as clearly defined as it is with the NP complements dis-

KUTI/+indic/

KUTI/“ subj/

KANA/-fpartic/

KANA/+coasec/

KU^7+infin/

ZVA,7+partic/

JOJTI _ ^+indic/

KUTI/Tsubj/

+ + +

KANA /+par tic,7 + + +

KAMA/+coxisec/

+ +

KU/+irtfin/

+

ZVA^/+part£c/ + + +

288

cussed in 6*1 above* While it is normal to talk of the grammatical fun­

ction of a noun phrase complement in various positions, it is not clear what the grammatical function of a verb phrase complement is*

sion can complement two types of verbs. First, they can occur as comple­ments of verbs which are expressive of modality or adverbiality. Secondly, there are those which show the extent to which the action (or state) of the higher verb is carried out (or entered into). I discuss each of

these two types in turn*

6.2*1 Verb Phrase Complements of the first typeThese occur in a structure of the type shown below in which

S is the complement in question*.

In the examples below, as well as in the. rest of this section, the com­plement clauses in question are underlined*

A. Complement clauses introduced by kuti / + subjunctive/

Broadly speaking, complement clauses of the type under diseus-

S

V S

, . i t / / f 4 / t / / / / / / /(21A) Chibage chinofanira kuti chixswe fetereza yepamusoro mazuvafano.(The maize crop should be top-dressed with fertilizer thesedays.)

(2 15) Fosa ndiye akasaka kuti ndiende^ kudhorobha* (Kosa was the one who made me go to town.)

B* Complement clauses introduced by ku £+ infinitive/

(216) Chibage chinofanira kuiswa fetereza yepamusoro mazuva ano (see 21h)

(217) Sekuru achanguno kudya*

289

(Uncle is still having his meal*)/ / /(2 18) Mangwana tichafuma kunorima mudoro*

(Tomorrow we shall go early in the morning to plough in thegarden.)

f i t / / / /(219) Vana vanogona kunzi vauye neChipiri*(The children can he made to come on Tuesday*)

In the following sets of examples the complementizer formative is obli­gatorily deleted (see also 2,8).

C. Complement clauses introduced by kana / + participial/t i f i / / / / / / / /(220) Vana vakaramba vachitaura vanzi naticha nyararai*

(The children kept on talking after they had been told by the teacher to keep quiet*)

(221) Ndinogara ndicliienda kumusha mwedzi'yose kumoonsf amai.(I go home every month to see my mother.), , m / n . /, /. _ , / / / . / ' ' , , / / y(222) Tull akachitemiwa chematsenganzungu akabva aparuma pasi.(Tuli was given a hard blow on the cheek and fell down (lit.bite the ground))

D. Complement clauses introduced by kana / + consecutive// f t I f f(223) Wagona ukabva wasvika, tanga toshupika.

(You did well to arrive now, we were getting worried.). . / / / / / / /(224) Vana vaita vakarega kuyambuka mhiri nhasi vainge vadziviri-

/rwako nemvura.(It's a mercy the children didn't go over the river, other­wise today they would have been cut off by the floods.)

&..2.1.1 Modal-like verbsWith the meaning that the verbs in the higher sentences express

in these sentences, the verbs in (21 -) - (22^) must occur with a comple­ment clause*. Another pertinent observation is that these verbs do notseem to be true verbs semantically. Some of them indicate modality as

r \ t * fm (21Zf) -fanira (must, ought) and -gona (can), in (219).. Semantically

290

these compare very well with the English, modais can, must, shall, haveetc.. Others express an adverbial idea of one kind or another as in (217)

/ / / /-nguno (do to this time, still); (218) -fuma (do early in the morning)-(220) -ramba (do continuously); (221) - g a r a (do often, always); (222)-bva (do then);: and (223) -gona (do well, it is a good thing that ) „Nevertheless, structurally they behave like any other verb. These adver­bial-like verbs require embedded sentences as their complements. Normally no material is permitted to intervene between them and their complement clauses. Adverbial enclitics like -su (expressing dissatisfaction), -wo (also) and -ka (form, of emphasis) may however occur with the higher verbin these sentences..

t i t / / t / t / / / / ^Chibage chinofanirasu kuiswa fetereza yepamusoro mazuva anoi

(ep 216)(Should the maize crop really be top-dressed with fertilizer these days?)

/ ✓ / / / / / / / / / / /Chibage chinofanirawo kuiswa fetereza yepamusoro mazuva ano.(The maize crop should also be top-dressed with fertilizer

these days,)t / / y / / / / // / / /Chibage chinofaniraka kuiswa fetereza yepamusoro mazuva ano.

(The maize crop should certainly be top-dressed: with fertili­

zer these days.)

Contracted Infinitives Included in this type of verbs are those which may be accompa­

nied by contracted infinitives.. In (228) - (231) the verbs are followed by full infinitives. But in (232) - (235) these same verbs are accompa­nied by contracted infinitives. The meaning of the sentences in (232) -(235) is the same as that of the corresponding sentences in (228) - (231)*

(228) Kunze kwanyanya kupisei/ kukadai?(Why is it so hot?)t i / / /(229) Jeke afuma kuenda kunotema map an go.

(225)

(226)

(227)

291

(Jack went to cut poles early in the morning.)✓ t f / t ✓ / /(2-30) Myatsa kudya zvako, pachime nguva.,

(Take your time to have your meal,.- there is still time.)/ V i 1 t / / / /■(231) Kana waenda kuchitoro uchimbidze kudzoka.

(When you go to the store, do come back early*)(232) Kumze k wanyanyopisei^kukadai?(233) Jeke afumoenda kumotema mtfpango.(234) Kyatsodya zvako, pachime mguva'.

/ / / / / / /(233) Kama waenda kuchitoro uchimbidzodzoka».With the verb -nguno (do still) the termiimal vowel never changes.

(236) Usamudaidze achangumo kuita basa ramdamutuma.(237) Usamudaidze achangumoita basa ramdamutuma..

(Don't call him, he is still doing the piece of work which I asked him to do.)

In the sentences above the infinitive contraction transforma­tion is optional. With other verbs however this transformation is obli­gatory as in the following sentences:

/ / / / / /(23o) a.. *Arara kuchema memanyoka.-/ / ✓ / f tb*. Arar0 chema memanyoka ..(She had stomach truubbe all through the might.)

(239)a. *Mashiri anogara kuregerera mombe dzake dzichxdya chibage chaSaru.

/ ✓ / / / ' /b. Mashiri anogaroregerera mombe dzake dzichidya chibagechaSaru.(Mashiri often lets his cattle, destroy Sam's maize crop.)

, . i / / / / / / / / /(2A0)a. *Sei muchiswera kutamba vamwe vachiita basa?t ✓ / / / ✓ / / / /b* Sei muchiswerotamba vamwe vachiita basa?

(Why do you spend the day playing while others are busy working?)

In the following pairs of sentences the uncontracted forms are even

292

more bizarre than, those in (238a), (239a) and (240a). However the embed­ded clauses, in the (b) sentences are no less contracted infinitives than those in (232) - (239) above,

(241)a. *Wamba/ kuona Sara her/ pano?i t / i f fb, Wamboona Sara here pano?

(Have you seen Sara here at all?)(242)a. *Zvaasina kuuya ndota kuenda kumusha ini*

b, Zvaasina kuuya ndotoemda kumusha m i .(Since he didn't come, I shall have to go home myself,)

(243)su *Ndimomda kutenga chii'kuchitoro ?/ / / / / /b, Ndimfrndotenga chii kuchitoro?

(What shall I go and buy from the store?)

6*2*2 Verb Phrase Complements of the second type:In the second type the verb in the higher sentence is not mece

ssarily accompanied by a complement clause in the same sense as was the case with the first type. The sets of sentences (244) ™ (2d?) and (248)- (231) illustrate this.

, . t / / / / ✓ / / /✓(244) Tii yacho yaiva yatonhora kuti vamhu vainwe.(The tea was too cold for people to drink it.)

f / i t /(243) Pasi paorna kuti murirae mumda uyu iye zvino.(The ground is too dry for you to plough this field now.)

(246) Hdakwana nenzara kuti ndiburitse ndima iyi.(I am too hungry to finish this piece of work.)

(247) Sabhuku uyu /tonga mh/sva dzakawanda kuti /tadze yako iyi*. (This headman has tried so many cases that he cannot failto try yours.)

The verbs in the higher sentences in (244) - (247) appear below withoutcomplement clauses.

f / / / / / /(248) Tii yacho yaiva yatonhora.

293

(The tea was cold.) t /(249) Pasi paoma.

(The ground has become dry.)(230) Ndakwana nenzara.

(I am very hungry.)/ N t * / ' t(231) Sabhuku uyu atonga mhosva dzakawanda,

(This headman has tried many cases.)Admittedly there is a slight difference in connotation between the verbs in the higher sentences in (244) - (247) and the corresponding verbs in(248) - (231). It is quite clear though that the verbs in these pairs of sentences are related to one another in the same manner. In (244) -

(247) it is implied that the action (or state) of the verb in the higher sentence is carried out (or entered into) to such an extent that some­thing happens or fails to happen. The complement clause fills in this latter part. The verbs in (248) - (251) on the other hand merely express the action (or state) without indicating the extent to which that action (or state) is done (or entered into).

One other noteworthy point is that the verb in the higher sentence may be stative or non-stative. By stative here is meant inchoa­tive, that is, verbs which are capable of expressing a present meaning although they may be in the past form* With stative verbs (and adjecti­ves also) the complement clause follows immediately as in (252) - (256) as well as in (244) - (245)* Ike element which signals the presence ofthe complement clause is not overtly expressed with these verbs.

, , / / / / / / /(252) Zviyo izvi zvawanda kuti zvikwane mudura umu.(This finger millet grain is too much for ijs all to be put into this granary,)

(253) Pesi akura kuti ave mubhuku rechina.(Pesi is too old to be in Grade Four.)

29^

/ V f t 1 t { / / / ✓(25A) Imiba yedu idikx kuti iitirwe. bira*.(Our house is too small for the bira to be performed in it*)

. / / / / / / /(295) Hdafa nenyota yedoro kuti ndimbosvika kumba ndigozodzokerakubhawa.(I am; too thirsty for beer for me to go home and then come back to the beer hall*)

(256) Aguta hwahwa kuti atyaire motokari.(He is too drunk to drive a car.)

But' with nom-stative verbs it is necessary to insert into the object noun phrase an adjective or a relative clause which expresses quantity before the complement clause in question is supplied. And if the non- stative verb is not followed by an objeat noun phrase, an adverbial of frequency is added before the complement clause is supplied. Sometimes even if there is an object noun phrase the adverbial of frequency repla1 ces the adjjec.tive or relative clause., of quantity.

(257)

(258)

(259)

(260)

If the element which expresses quantity or frequency is left out from •

t t i t ✓ / / / / / / / / ■Mari aona zvakawanda kuti atorwe mwoyo nevasikana vakaita/saSekai.

(Mari has seen too many things to be attracted to such girls as Sekai.)

/ < / / / / / / /Hdanzwa zvmhu zvizhinji muupenyu hwangu kuti nditirimuke/ / /zvauri kutaura*

(I have heard too many things in my life to be affected bywhat you are saying.)

/ / / / / / / / t s ✓ f s * sSekuru vakauya kuno kuDomboshava kazhinji chaizvo kuti vataike.(Uncle came here to Domboshava so many times that he won't get lost.)

Mvura yapenya chaizvo manheru ano kuti irege kunaya usiku. (There was so much lightning this evening that it can't fail to rain during the night.)

a95

(257) - (260) the resulting sentences are either ungrammatical or bear a different meaning altogether,

(261) *Mari aona zvinhu kuti atorwe mwoyo nevasikana vakaitasaSekai.(*Mari has seen things for' him to be attracted tosuch girls as Sekai.)

/ v f / / / / / / f(262) *Ndanzwa zvinhu muupenyu hwangu kuti nditirimuke zvaurikutaura. (*I have heard things in my life to be affectedby what you are saying,.)

, N / / / / / / / / / / / /(263) * Sekuru vakauya kumo kuDomibo shava kuti vataike,(■*Umele came here to Domiboshava to get lost,)

/ N / / / / / / / / / / *(26 -) *Mvura yapenya manheru ano kuti irege kunaya usiku.(*There was lightning this evening that it can’t fail to rain during the night.)

These sentences are ungrammatical if in plac.e of the adjective or rela­tive clause of quantity we put in any other qualifiers.

i i ( i t / / / / / / /(265) *Man aona zvinhu zvikuru kuti atorwe rnwoyo mevasikanavakaita saSekai.(*Mari has seen great things to be attracted to such girls as Sekai.)

/ i t t / / / /(266) *Ndanzwa zvinhu zvinonakidza muupenyu hwangu kuti nditiri-t f /mruke zvauri kutaura.

(*I have heard interesting things in my life to be affected by what you are saying*)

6.3 SummaryIn this chapter I have tried to show the grammatical distri­

bution, or function, in sentences of complement clauses, both those that are dominated by NP and those dominated directly by VP. From the account given above it is evident that the role of complement clauses in Shona cannot be minimized. We see them function as subjects, objects and also as adverbials of one kind or another.

296

7.0 SOME RESIDUAL PROBLEMSThe account given in the foregoing chapters of sentential com­

plements in Shona is by no means exhaustive. There are many areas whichhave as yet to be investigated. Some of these areas are outlined brieflybelow.

7*1 Relationship between verbs and ideophonesThe relationship between verbs and ideophones in sentences

like the following requires further investigation. Both the ideophonesand the complement clauses are underlined.

/ / / / / /(1) Chimuti chakati pwa kutyoka.(The stick got broken.)

, . // / / / / / /(2) Romi anogara akati zii kunyarara.(Roni is always quiet.)

f t / ' /(3) Damba rakati bu kuwa mumuti.(A wild orange fell down from a tree.)

/ / / /The subjects of the infinitives kutyoka in (1), kunyarara in (2) andkuwa in (3) are clearly chimuti, Roni/ and damba respectively. Therefore it is uncontroversial that these three infinitives are embedded clauses of some sort. It is interesting to note that the meanings of these in­finitive verbs and that of the preceding ideophones are the same. These explanatory infinitives, as I may call them, are added to the sentences so as to make explicit the meanings of the ideophones, for it is not infrequent for an ideophone to express more than one meaning as in (1) and (3).

/X / / / / /( )a* Mvura yakati pwa kutinhira.(There was lightning and great thunder.)

/ / / / ✓ /b* Rumuti rwakati pwa kutyoka.(A stick got broken.)

/ / / / /(3)a. Mombe ndakaiti dzi Bedemo kutema.(I cut the ox with an axe.)

29?/ / /b. Nyatsoti dzi kumira* '

(Stand upright*)In cases where the meaning of an, ideophome is unambiguously clear, either because the ideophome has only one shade of meaning or from the context, this explanatory infinitive need not be present*

These explanatory infinitives' may also accompany other verbs as in (6) below*

t J / / /(6)a* Sadza mdakarirakasha km.dya.(I ate a lot of sadza.)

/ / / / / / / / /b* Vana vakadyisa mumunda rnangu ndakavarakasha kurova*(As for the children who let their animals destroy my fieldI thrashed them severely.)

/ / / / / * / /c. Chibage ndakachirakasha kutema*(As for the maize crop I cut down a lot of it.)

The verb -rakasha in (6) carries a slightly different connotation in each sentence as is indicated by the explanatory infinitives which are underlined*

Notice also that if the preceding verb is in the passive, the verb in the explanatory infinitive will of necessity be in the passiveform also*,

, * / / X / / x(7)a. Rumuti rwakamzi pwa kutyorwa*(The stick was broken.)

t / * f , t /b*. * Rumuti rwakamzi pwa kutyora./ON / ^ / / / /(o)a* Mombe yakanzi dzi medemo kutemwa.

(The ox was cut with an axe.)/ % ' / / /b. *Mombe yakanzi dzi medemo kutema.

(9)a. Chibage^ chakarakashwa neni^kutemwa.(The maize crop was hewn down by me.)

/ / / / / / //b* *Chibage chakarakashwa neni kutema*Observe also that complement clauses which occur after ideo-

298

phones are restricted to those which are introduced by ku /+ infinitive./only as the examples in (1) - (9) above show.

However, what is not immediately clear to me is how to charac­terize these explanatory infinitives in terms of deep and surface struc­ture. That is, should they bs; regarded (a) as ideophone complements^’u in which case they would appear in structures like the following:

S

NP VP

Ideophone Phrase

Ideophone S

in which S will give rise to the explanatory infinitive;or (b) as verb phrase complements

NP VP

V Ideophone Phrase S

or (c) as sentence complements

NP VP S

V Ideophone Phrase

or (d) even as a conjunct of some kind.

I 299

S

SS.

NP VP

IdeophonePhrase

Since the ideophone and the explanatory infinitive complement each other it would appear it is preferable, to treat the explanatory infinitive as a verb phrase complement. This would make them sister nodes. It must be emphasized though that I am merely stating a preference rather than making a statement of fact.

Verbalized ideophones may still be accompanied by explanatoryinfinitives in the same way that plain ideophones are.

. . / / / ✓ / /CIO) Peni akapwatika kuseka.i f f / /cp* Peni akati pwati kuseka.

(Peni burst out laughing.), x / / / / / / / / / /(11) Nyoka yakamukon1ora kuruma pagumbo.

/ / / / / $ / / /cp. Nyoka yakamuti kon* o kuruma pagumbo.(The snake struck him on the leg.)

(12) Ruzvidzo akabvotorooka kutaura padare,✓ ✓ ✓ / /cp. Ruzvidzo akati bvotomo kutaura padare.

(Ruzvidzo spoke unwisely at the court.)The pairs of sentences in (10) - (12) seemi to be related both structur­ally and semantically. But it has still to be determined precisely in what way they are related.

Notice also that if the verb in the explanatory infinitive is interchanged with the verbalized ideophone the resulting string is un­grammatical.

300

(13) *Peni/ aseka kupwatika'. (cp* 10)/ / /(lif) *Nyoka yakamuruma kukon' ora*. (cp.. 11)

(13) *Huz.vldzo akataura kubvotoiraoka padare. (cp. 12)/ / /This applies equally to verbs like, -rakasha which readily accept expla­

natory infinitives.,/ i f / /(16) a. #Set&za ndakaridya kurakasha. (cp.- 6a)

b.. * V. 8lXlsl vakadyisa mumunda mangu. ndakavarova kurakasha.. (cp.. 6b)c. *Chibaga’ ndakachitema kurakasha. (cp.. 6c)

7*2 An aspect of the possessive constructionTbe following constructions are noun phrases which consist

each of a head noun and a possessive,(1?) badza/ rekusakurisa (a hoe with which to weed)(18) mar yekufambisa^ (money with which to travel)

The possessive construction in these noun phrases comprises a possessive morpheme and an infinitive verb*. The infinitives in rekusakurisa and/ / s tyekufambisa are no doubt embedded clauses. That they are embedded clau-

ses becomes more obvious in sentences like the following:(19) Toko* anoda/ badza rekusakurisa.

(Toko wants a hoe with which to weed (the field).)/ N / s / / / / s /(20) Bere akapa rnwana mari yekufambisa.

(Bere gave the child some money for it to use on. the journey,.) The. grammatical subjects of the infinitive constructions kusakurisa and kufambisa' in (19) and (20) are respectively Toko''' and mwanal The diffic­ulty arises in how to characterize the noun phrases in (17) and (18)* Precisely how is the possessive element, represented in these noun phra­ses. by /e/, as well as the causative idea represented by the morpheme /is'/, to be described in terms of deep structure? As a possible solution the possessive element may be generated by the phrase structure rules, and to account for the causative idea one might have to postulate an ah-

301

qstract. verb. along the lines suggested, by G*iako£f* For the moment I

leave it as an open question as to how to describe the type of noun phrase in (1?) and (l8)*>

7*3 Complement clauses introduead by kuti AsuhJ/and ku. A iELfiB./Another problem relates to complement clauses introduced by

kuti /+ subjunctive/* and ku /+ infinitive/.- We saw in the preceding chapters that some complement clauses introduced by kuti /+ subjunctive/ were synonymous with those introduced by ku /+ infinitive/* Considerthe following pairs of sentences for illustration*.

/ / / / / / / /(21)a. Dende anoda kuti atenge motokari*/ i t / / /b. Dende anoda kutenga motokari*.

(Dende wants to buy a car*.)t i t t i t / / / < ✓(22) a. Hyoka yac.ho akaedzesera kuti airove ikabva yamusvipxra.,t / / t / / / ✓ n / / X /b* Nyoka yacho akaedzesera kuxrova ikabva yamusvxpira*

(As he tried to hit the snake it spat'.- at him.)/ / / t / / / /(23)a*. Tondi abvurna kuti adzokere kumo hwedza.t i t / t fb. Tondi abvumia kudzokera kuno hwedza*

(Tondi has agreed to return here tomorrow.)* , • * * ' *(2A)a. Unogona kuti uuye nenguva yaunoda*.✓ i t / t , /b. Unogona kuuya nenguva yaunoda*

(You can come at the time you want*)/ / J / S / .(23)a., Varangana kuti vatize muqeri*/ * . y / * / .b* Varangana kutxza mujerx*

(They have planned to escape from prison.)If there were a way of accounting for the synonymy of the sentences in the pairs (21) - (25) using a transformational rule(s), our grammar would be; simplified*. Since the difference between the members of each

1*. George Lakoff in Irregularity in Syntax (1970);; see also Robin T*. Lakoff in Abstract Syntax and Latin Complementation (1968),.

302

pair lies in the complement clauses only, namely, that one is introduced by kuti £ + subjunctisre7 and the other by ku / + infinitive/, a solution would be to take one of the two complementizers as generated by the? phrase structure rules and the other as. introduced by an optional trans­formational rule under a certain stated condition, namely, that the subject of the complement clause must be identical with that of the high er sentence*. Two objections to this solution immediately arise. First, these two complementizers are not always substitutable one for the otheras the pairs of sentences below show*

, / / / / / / f / ✓ / /(2b) a. Rozi akaita kuti aroorwe. chete naHund*(Rose worked it in such a way as to get herself married by Huni.)

///' i / / / / /b. *Rozi akaita kuroorwa chete naHuni*./ N / / / / ✓ ✓ / / / / /(2?)a* Foromani akapunyaira kuti aone zvekuita nenyaya yekudzingwa

/ / /kwake basa*(The foreman thought very hard about what he could do con­cerning his dismissal from work.)

i / / / / / ✓ / / i f f /b* *Foromani akapunyaira kuona zvekuita nenyaya yekudzingwakwake* basa*

(28)a. *Mwana wako abata kuti abel b* Mwana wako abata kubal

(Your child has started stealing.)/ / / /(29)a* *Anyarara kuti acheme./ / /b* Anyarara kuchernia*(She has stopped crying.,)/ / / i f(30)a. *Vapedza kuti vataure zvakaitika*

b*. Vapedza kutaura zvakai/bika..(They have finished narrating what happened.)

Given the transformational rule suggested above, (26b) and (27b) would

be grammatical sentences, which they are not. There would be need then

303

for a caveat to this rule so that it excludes (26b) and (27b)* Further­more, since (28a), (29a) and (30a) are ill-formed, there would be need also to vary our transformational rule above so that it applies, not optionally, but obligatorily in these cases. With all these caveats it is doubtful whether in the end our grammar will be any simpler* Secondly, introducing ku /+ infinitive^ transformationally seriously undermines the hypothesis that complementizers are present in deep strrc ture which was otherwise convincingly argued for in Chapter Four, In any case it is highly questionable, if some complementizers are genera­ted by the phrase structure rules while others are introduced transfor­mationally, whether our grammar will achieve any reasonable level of descriptive adequacy at all. At the moment I know of no way of relating structurally within the generative transformational model the members of the pairs in (21) - (23)*

30 4

8*0. CONCLUSIONIn this study I have endeavoured to show the following points;

firstly, it clear that an adequate linguistic description of Shona should take account of the fact that some embedded sentences, are- domi­nated by NP, while, others are dominated directly by VP*, Given the sent­ences

, . r s / i t /(1) Vanhu vazhinjr vanotemdera kuti Mwan ariko*(Many people believe that there is God.)

(2) Kuti Mwari^ ariko* zvinotenderwa nevanhu vazhinji.(That God exists is believed by miany people.)

t / /it is as easy to believe that the clause kuti Mwari ariko is an instanceof the constituent S as it is to believe that it is an instance of theconstituent NP. A linguistic, description which does not postulate that

/ t /the string kuti Mwarih ariko is an S at some level of derivation cannot achieve empirical adequacy. However, taken in the context of (1) this string is functioning grammatically as a direct object noun phrase, hence passivization. is possible as (2) illustrates. This string is a noun phrase in (1) because it passes the tests for noun phrases and it behaves like an NP. We are therefore able to say that a given complement clause belongs to one or the other of these two groups, that is NP com­plements and VP complements, by applying the tests for discovering noun phrases. It is claimed in this study that any complement clause which passes these tests is considered to be dominated by NP, while those which do not pass these tests are taken to be dominated directly by VP* One advantage of this approach is that once the tests for discovering simple noun phrases have been determined and established these same tests can be applied also to discover long and complex noun phrases* As has already been observed during the course of this discussion not all tests for noun phrases apply to every noun phrase in any given sentence.

305

For instan.ce, noun phrases which function as adverbials have fewer tests applicable to them1 than noun phrases which are in an object position.

Besides, it has also been demonstrated conclusively that it is necessary to posit a head noun of either class 8 or class 15 (or class 16 in a few cases) for complement sentences which are dominated immedia­tely by NP in. order to account for the concordial agreement which these complement sentences govern. The other alternative, namely, to regard such an NP as expansible into B only has been shown to result in a very weak grammar.. As I have already pointed out the position with regard to the complement clauses which are directly dominated by VP is not so clearly defined* With further investigation it may turn out that the verbs discussed in 6.2*1 would be better treated as modals, and not as verbs as such, although this is not immediately apparent to me. The sec­ond type of VP complements may also prove not to be a watertight group­ing. For one thing they occur in sentences which express comparison.With further research into how comparison is achieved in Shona this type of VP complement may turn out to require a different characteriza­tion. For the purpose in hand however any complement clause which does not meet the tests for noun phrases is considered to be a VP complement..

It has also been shown that a linguistic description of Shona should include complementizers among i#s grammatical categories. Comple­mentizers have a semantic, content although sometimes this meaning is not easily and immediately extractable. Accordingly in any derivation which involves a complement clause complementizers will be specified in the deep structure, and not introduced into sentences transformationally. Furthermore, the grammatical category of complementizer (COMP) is taken to comprise two elements, namely, a complementizer formative and a mood. This study claims also that every sentence node starts as S, implying that every sentence has a complementizer. In the case of the topmost S

306

the complementizer formative will be deleted obligatorily by a conven­tion. It has also been shown that complementizers subcategorize verbs and also that these complementizers are mutually exclusive.

Another point which emerged in this study is that a descript­ion of Shona will be more adequate if we adopt the transformational approach for deriving infinitival constructions and the lexicalist approach for the other derived nominals. While derived nominals are always dominated by N3?9 infinitives on the other hand may be dominated directly by either NP or VP.

307

APPENDIX A

Phrase Structure Rules for Shona n, 0 _________ v (PreS) NP + AUX + VP (PP) (ADV)11JL * o

C Imrp.'l C Passive ^2*. PreS ---- ^ ( ^Q j ) (leg*) ( ^Pseudo-passive| )

s f f s “ " ] ) !3. Vp J J V(COP + Pred* /

h* Pred*. -- ^ /lP\ PP Adj..Demi*Sel*Quant *Poss*

3* AUX ----- £ Tns (Aspect)6*. PP --- > P + NP7. KP -------> C * (S) 1

Vnp + s J8. S — ^ COMP + S9* COMP ^ /'kuti 7+ indicative/, kuti /+ subjunctive/,

kana /+ participial/, kana /+ consecutive/, ku /+ infinitive/, z.va /+ participial/, vi complementizer j

10* ADV*'---— ^ (Adv..) (Time) (Place) (Manner) (Frequency)11* N -----— ^ (mukomiana, miapanga, upfu----- )12* V ----- •v/ / * * * * * \— 7 (-rova, -nyora, -fantba----- )13. Tns* '— -— ^ (a, ai, aka, cha, nga, no)1A-. Aspect— — ^ chi15. P — — _ — ^ (na, sa, ir )16. A&J* — r ? * f ^— 7 (-kuru, “tema, -tsva --)17. Demi* — ^ ( chi, va, k a -----)18. Sel* --— ^ (-no, -ye )

308

*

*

>

Abbreviations:A $3*. — Adjective;AUX. = Auxiliary;Dem.. = Demonstrative; Neg*, = Negative;:Poss. - Possessive; PP = PreS = PreSentence;S = Sentence;V = Verb;

19. Quant.,20* Adv. —21. Time22. Place —— -23. Manner --21+m Frequency

-o se, ~@ga)/ /chete, bedzi, kani ---- )

nhasi, lezuro, mangwana, ---- )

immba, kurukova, panzira----- )chikomana, zvishoma, zvinonwirwa mvura — kaviri , rutatu^,----- )

ADV. =2 Adverbial; Adv. = Adverb;COMP = Complementizer; COP. = Copula;Imp. = Imperative; If = Noun;NP = Noun. Phrase; P = Preposition;Prepositional Phrase; Pred* ~ Predicate;Q = Question; Quant. = Quantifier;Sel* = Selector; Tns = Tense;VP — Verb Phrase;

309-

appendix B

The. verbs which are listed below are grouped according to the complementizer or complementizers which they allow in their comple mient clauses*,

1*. Verbs which allow kuti /+ subjunctive/ and ku /+ infinitive/ in their complement clauses:

“bvurnia (agree) * * *-chimbidza (hasten) ~da (want, need)// /“dziidza (learn) -edza (try) -edzesera (attempt.)

-funda (learn) -funga (think) "fungidzira (guess)"gadzirira (be. prepared) “gadzira (prepare) -gara (not to do)-gona (manage) / / / .-jaira (be in habit of) / / / .-kanganisa (do wrong)-kanganwa (forget) "kodzera (be worthy) -kona (fail)-koshiwa (forget) / ✓ s-kungura (regret) / f /-kwanisa (be able)-mirsf (stop) -matsa (do well) -netsa ( give trouble)-ngwara (b.e wise) -nonoka (be late) “nyangadza (spoil)-nyanya (do excessively) -ayar/ (be ashamed) -nyunyuta (grumble)-nzvenga (avoid, dodge) -nzwa^ (hear) -raraba (refuse)-rangana (plot, discuss) -sema (detest) “Shinga (exert oneself)“shora (despise) -siya (leave) / /-tadza (fail to do)-t/ta (be scared) -tiza"(run away) “tsika-tsika (hesitate)-tsutsumwa (be reluctant) -tsvaga (seek) /“tya (fear, be afraid)-udza (tell) -vamba (begin) * / S f ^-vavarira (persevere)-z.engurira (be unwilling -zvonda (hate)

to exert oneself)

2*. Verbs which allow kuti / + indicative/ in their complement clauses:-bheja (make solemn

promise)/ / / v-bvotomoka (speak unwisely)

-bvuma (agree) -bvunza (ask) / / t“Cherechedza (observe)

310

■daira (believe)■funga (think)kanganisa (do wrong)■natsa (do well)■nzwa (hear)•pfidza (confess)■pokana (refute)■rangarira (ponder)

/ f■semn (detest)■tadza (fail to do)■temda (thank)■tsanangura (explain)

/>tya (fear)■viga (hide)

-dziidza (learn)-fungidzira (think)

/ / /-kanganwa (forget)i /-nyara (be shy)

-oia (see)-pika (swear)-ramba (refuse)

' / r \-reva (mean)-shinga (exert oneself) -tangs (begin)-tendera (believe) -tsiidza (promise) -vamba (begin)-yeuka (remember)

-fembera (guess)/ / /-fushunura (reveal)

-koshiwa (forget) -nyunyuta^ (grumble)

t / / .-paridza (preach)-pindura (reply)-rangana (plot, discuss) -ripa (compensate for) -shora (despise)-taura (speak)

/ f /-tondera (remember)-tsiftsumwsf (be reluctant) -vaaza (hide)-ziva (know)

3» Verbs which allow kuti /+ subjunctive/ and kuti /+ indicative/ in their complement clauses:

-bvuma (agree)t i / f-kanganisa (do wrong)

-natsa (do well) -nzwa'(hear)-reva (mean)-shora (despise) -taura (speak) -tondera (remember)

* s ,-udza (tell)

-funga (think)/ ✓ t-kanganwa (forget)

-nyara (be shy)-ramba (refuse)-serna (detest)

/ /-tadza (fail to do) -tenda (thank)-1 su tsumwa (grumble) -vamba (begin)

-fungidzira (think)-koshiwa (forget)

/ / /-nyunyuta (grumble)-rangana (plot)-shinga (exert oneself) -tanga (begin)-tendera (believe)-tjf/ (fear)

Zf*. Verbs which allow kuti + indirect question as direct object:/ / / / / /-bvumza (ask) -cherechedza (observe) -dziidza (learn)

311

-fembera (guess)/ f / , s-fushunura (reveal)

-koshiwa (forget) -nz.wa/ (hear)-rangana (discuss)

f t . .-tadza (fail to do)-yeuka (remember ) -tsanangura (explain)

-funda (learn) -gadzira (prepare) -nanganidza (look at) -ona (see)-rangarira (meditate) -tsvaga (find)-ziva* (know)

-funga (think)-kanganwa (forget)

/ * /-nyunyuta (grumble)/ / / ( ,-ongorora (examine)

-shaya (be at a loss) -verenga (count) -pi nyaira (be worried)

3*. Verbs which allow ku /+ infinitive/, but not kuti /+ subjunctive/ or or kuti /+ indicative/} in their complement clauses:

(a) Infinitives accompanied by possessives-bata (get usec|to) -chinja (change) -enzana (be alike)

/ /-gura (give up bad habit)-nyanya (do excessively)

/ ✓■isva (do badly)/ * 6 f V-kurumidza (hurry)

-nyarara (be quiet) -pamhidza (do again) -pora (be cool)■serera (subside) -wanza (do many times)

(b) Infinitives unaccompanied by possessives / /-bata (get used to) -enzama (be alike)

i f i f-isva (do wrong) -kwana (be affected)/ i-nyanya (do excessively)-nyarara (be quiet)

/ /-pedza (finish.) -pinda (get used to)-serera (subside) -wanza (do many times)

i f , ,-gura (give up bad habit)f / /-kurumidza (hurry)

-pamhidza (do again)/ / ,-pora (be cool)

6. Verbs which allow kuti /+ subjunctive7? but not ku /+ infinitive/,in their complement clauses:

-ita (do)/ i-sunga (decide at court)

i f * , ^-paridza (preach) , / / ,-terna (decide at court) / i *-punyaira (be worried)

312

7* Verbs which allow kuti, /+ subjunctive/ in their1 complement clauses in which the subject may or may not be identical with the one in the. higher sentence:

-da (want, need)-fungidzira (think)

-bvuma (agree) -fumga (think)

-edza (try)

■gadzirira (prepare,) -goma (manage)/ / t-kanganisa (do wrong)

-natsa (do well)t / . .-nyara (be shy)

-ramba (refuse)

-kanganwa (forget) -netsa (be difficult) -nyunyuta (grumble)■rangana (plot* discuss) -reva (mean)

-gadzira (prepare)/ f S , V-jaira (be in habit of)

-koshiwa (forget)I I I ,-nyangadza (make mistake)

/-nzwa (hear)/ /

-shinga (exert oneself) -shora (contemn)/ i t-tsutsumwa (be reluctant)

■tya (fear) i t , * , x■vavarira (persevere)

/ /-tadza (fail to do)-tsvaga (seek, find) -zvonda (hate)

8.. Verbs which allow kuti subjunctive/ in their complement clauses in which the subject may only be identical with the one in thehigher sentence:/ f-chimhidza (hurry)

-funda (1earn)/ / / f-kungura (regret)/ / /-shena-shena (do energetically)

/ /-tanga (begin)

/ / /■ dziidza (learn)-gara (not to do) -mira (stop)

-edzesera (attempt) -kodzera (be fitting) -nonoka (be late) -siya (leave)

-tsika-tsika (hesitate) -varaiba (begin) -zengurira (be unwilling to exert oneself) -nanga (go straight)

9* Verbs which allow kuti /+ subjunctive/ in their complement clauses in which the subject may not be identical to the one in the higher sentence:

/ t / * /-ita (do) -konzera (cause) -paridza (preach)-tema (decide) -sunga (decide at court) -tonga (judge)

313

BIBLIOGRAPHY Bach, Emmom (196if)

Am Introduction to Transformational Grammar. New York: Holt, Rinehart and Wins tom, Inc..

(1968)"Nouns and Noun pirases" in E.Bach and R.Harms (eds.),Universals in Linguistic. Theory.. Hew York: Holt, Rinehart and Winston, Inc,

Bresmam, Joam W, (1970)"On Complementizers: Toward a Syntactic Theory of Comple­ment Types", Foundations of Language 6, 29? - 321

Chomsky, N, (19 37)Syntactic Structures, The Hague: Moutom.

____________ (1963)Aspects of the Theory of Syntax. Cambridge, Mass.: M.I.T* Press,

____________ (1968a)"Remarks on Nomdmalizatiom" in R.A.Jacobs and P.S.Rosen­baum (eds.,) , Readings in English Transformational Grammar. Waltham, Mass.: Ginn/Blais dell*.

_____________ (1968b)"Deep Structure, Surface Structure and Semantic Inter­pretation" in D.D.Steinberg and L.A.Jakobovits (eds.), Semantics: An Interdisciplinary Reader in Philosophy, Linguistics and Psychology. Cambridge University Press.

Dembetembe, N.C. (1969)Verbal Constructions in Korekore. Unpublished disserta­tion for M.Phil. , London: University of London.

d'kovv\<; k\. /V. (jHLUcl)FaJilm©-re-,— C.-J-*— (-19-68-9 „ . , ,. ,lL Jke . L o9 (W l BotSi’s- oi Li^gn<V*c‘c t j V o ^ s 09 Hv f. C- c t-vgr e-SS ^ ■ Tke. H <

cuu— S’vs v

31^-

C ' T . (i*i<>$)

’’The Case for Case” in E.Bach and R.Harms (eds.), Universals in Linguistic Theory. Mew York: Holt, Rinehart and Winston, Inc,

Fortune, G. (1935)An Analytical Grammar of Shona. Cape Town: Longmans.

____ (Forthcoming)Shona Grammatical Constructions, Parts I and II. Salisbury: University of Rhodesia.

Fraser, Bruce (1970)’’Some Remarks on the Action Nominal!nation in English” in R,A.Jacobs and P.S.Rosenbaum (eds,), Readings in English Transformational Grammar. Waltham:, Mass.: Ginn and Company.

/Givom, Talmy (1969)’’Studies in ChiBemba and Bantu Grammar,” Studies; in- African Linguistics, Vol.. 3> Supplement 3* June 1972.

Gleason, H.A.. (1961)An Introduction to Descriptive Linguistics. Mew York: Holt, Rinehart and Winston, Inc.

Gleitman, Lila R. (1965)’’Coordinating Conjunctions in English”, Language, Al*260 - 293*

Gregersen, E.A. (1967)’’Prefix and Pronoun in Bantu”, Memoir 21, Supplement to International Journal of American Linguistics. Vol. 33j Number 3j July 1967*

Grinder, John (1972)”0m the Cycle in Syntax" in John P.- Kimball (ed,),Syntax and Semantics Volume I. New York: Seminar Press.

3X5

Grinder ,

Guthrie;,

Hankamor

Hasegawa

Heny, F.

Jacobs,

Katz,, J.

Kimball,

Kiparsky

John and Elgin, Suzette H. (1973)Guide to Transformational Grammar. Few York: Holt., Rine­hart and Winston, Inc.

Malcolm. (1948b)'•Gender, Fumber and Person in Bantu Languages," Bulletijfc of th.e School of Oriental and African. Studies. (1948) Vol. XII, Parts 3 and 4 (p.847 - 856)

, Jorge (1972)"On the Fonexistenee of Mirror Image Rules in Syntax" in John P .Kimball (ed„), Syntax and Semantics Volume I.Few York: Seminar Press,

f K., (1968)"The Passive Construction in English",; Language, 44; 2: 230 - 243-

(1972)"Bantu Lexical Classes and Semantic Universals", Studies, in African Linguistics. Vol., 3> Number 2, August 1972.

3 .A. and Rosenbaum, P..S.. (1968)English Transformational Grammar. London: Ginn and Co..

I. and Postal, P.M., (1964)An Integrated Theory of Linguistic Descriptions.. Cambridge, Mass,..; M.I.T. Press..

John P., (1972)"Cyclic and Linear Grammars" in John P .Kimball (e.d,.), Syntax and Semantics Volume I.. Few York: Seminar Press.

, P. and Kiparsky, C. (1970)"Fact" in M.Bierwisch and K.Heidolp (eds.), Progress in Linguistics. The. Hague: Mouton.

316

Koutsomdas, Andreas (1972)"The Strict Order Fallacy", Language. Vol.. 1 $ , Part I, P..88 - 96.

Lakoff, George (1968a)"Deep and Surface Grammar." Harvard University, mimeo- grphed.

_______ (1970)Irregularity in Syntax. Hew York: Holt, Rinehart and Winston, Inc*

Lakoff, G. and Peters, S. (1969)"Phrasal Conjunction and Symmetric Predicates" in D.A. Reibel and S.A.Schare (eds.*), Modern Studies in English. Englewood Cliffs, M.JV: Prentice-Hall, Inc..

Lakoff, Robin T. (1968)Abstract Syntax and Latin Complementation-.. Cambridge, Mass..: M.I.T. Press,

Langacker, Ronald V/.. (1966)"On Pronominal!zatioel and the Chain of Command" in D..A* Reibel and S.A.Scheme (eds.),- Modern Studies in English. Englewood Cliffs, H.J.: Prentice-Hall, Inc*

_____________(1969)"Mirror Image Rules I: Syntax", Language Zf5* 373 - 398.

Lees, R*B* (1960a)The Grammar of English Hominalizations* Bloomington,. Indiana: Indiana University Press.

Lees, R.B. and Klima, E.S. (1963)"Rules for English Pronominalization", Language* 39* 17- 28.

317

Marconnes, Rev.,F.> (1931)A Grammar of Central Karanga. Johannesburg: Witwatersrand university Press*

McCawley, 1*. (1968)"Concerning the Base Component of a Transformational Grammar", Foundations of Language. (4;:3: 2A-3 ~ 269)*

Par tee, Barbara. Hall (1971)"On the. Requirement that Transformations Preserve1 Mean­ing" in. C.J,.Fillmore and D.T. Langendoen (eds*), Studies in Linguistic Semantics. Hew York: Holt, Rinehart and Winston, Inc*

Postal, P.M. (196 -b)"Limitations of Phrase Structure Grammars" in J.J.Kata and J.A.Fodor (eds*.), The Structure of Language. Engle­wood Cliffs, N.J.: Prentice- Hall, Inc*

____________ (1966b)"On So-called Pronouns in English" in R,.A.Jaeobs and P.S.Rosenbaum (eds*), Readings in English Transformatio­nal Grammar. Waltham, Mass.: Ginn and Company,

_____________(1971a)Cross-Over Phenomena. H'ew York: Holt, Rinehart and Winston, Inc.

Ringem, Catherine (1972)"On Arguments for Rule Ordering", Foundations of Language. 8. (1972) 266 - 273#

Rosenbaum, P.S*. (1967)The Grammar of English Predicate Complement Constructions*.Cambridge, Mass,: M.I.T* Press.

318

______ (1970)"A Principle Governing Deletion in English. Sentential Complementation" in R.A.Jacobs and P.S.Rosenbaum (eds*.), Readings in English Transformational Grammar* Waltham, Mass*: Ginn and Company*

Ross, J*R* (1966a)"A Proposed Rule, of Tree-Pruning" in D*A*,Reibel and S.A* Schame (eds*), Modern Studies in English. Englewood Cliffs, N*J.: Prentice-Hall, Inc*_ (1966)"Adjectives as Noun Phrases" in D*A*Reibel and S.A. Schane (eds*), Modern Studies in English* Englewood Cliffs, N*J*: Prentice-Hall, Inc,

_______ _ _ (1966b)"On the Cyclic. Mature of English Pronominalization" in D*A*Reibel and S„A*,Schane (eds*), Modern Studies in English* Englewood Cliffs, N.J.: Prentice-Hall, Inc*

____________ (19670,)Constraints on Variables in Syntax* Unpublished doctoral dissertation. Cambridge, Mass*: M*I,T*.

Ruwet, Nicolas (1973)An Introduction to Generative Grammar. (translation by Norval S.H.Smith), Amsterdam: North-Holiand,

Stockwell, R*P*, Schachter, P, and Partee, B,Hall (1973)The Major Syntactic Structures of English, Mew York: Holt, Rinehart and Winston, Inc,

Voeltz, E* (1971)"Surface Constraints oit Agreement Resolution: Some Evi­dence fromr Xhosa", Studies in African. Linguistics. 2.1: 37 - 60*.